OPTION VALUES: Model directory: ./IOB2+0.02/ Training data file: train.tagged Testing data file: test.tagged Unlabeled data file: data.untagged Model file: model.txt Training log file (this one): trainlog.txt Label representation: IOB2 Second-order Markov CRFs Number of labels: 24 Number of training sequences: 443 (one data partition) Number of testing sequences: 22 (one data partition) Number of unlabeled sequences: 0 Number of context predicates: 690662 Number of features: 1471004 Feature rare threshold: 1 Context predicate rare threshold: 1 Using multiple rare thresholds for features: 0 Highlight feature: 0 Number of training iterations: 200 Initial lambda value: 0.0200 Sigma square (for smoothing): 100.0000 Epsilon for L-BFGS convergence: 0.000100 Number of approximated hessian matrixes: 7 Start to train ... Iteration: 1 Log-likelihood = -5521918.953409 Norm (log-likelihood gradient vector) = 542014.555707 Norm (lambda vector) = 24.256991 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 5327 4334 81.36 93.55 87.03 b-np 12220 11737 10096 86.02 82.62 84.28 i-np 13660 19252 13340 69.29 97.66 81.06 o 6349 5154 5038 97.75 79.35 87.59 b-vp 4768 3050 2972 97.44 62.33 76.03 i-vp 2602 1820 1790 98.35 68.79 80.96 b-adjp 384 21 21 100.00 5.47 10.37 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 64 63 98.44 7.66 14.22 i-advp 100 1 1 100.00 1.00 1.98 b-sbar 503 6 6 100.00 1.19 2.36 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 19 19 100.00 15.08 26.21 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 51.43 25.74 34.31 Avg2. 46451 46451 37680 81.12 81.12 81.12 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 11737 7605 64.80 62.23 63.49 pp 4633 5327 4331 81.30 93.48 86.97 vp 4768 3050 2686 88.07 56.33 68.71 sbar 503 6 6 100.00 1.19 2.36 adjp 384 21 21 100.00 5.47 10.37 advp 822 64 63 98.44 7.66 14.22 prt 126 19 19 100.00 15.08 26.21 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 63.26 24.15 34.95 Avg2. 23486 20224 14731 72.84 62.72 67.40 Current max chunk-based F1: 67.40 (iteration 1) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 2 Log-likelihood = -4986690.721297 Norm (log-likelihood gradient vector) = 526940.520225 Norm (lambda vector) = 24.315970 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 6710 4140 61.70 89.36 73.00 b-np 12220 10555 8516 80.68 69.69 74.78 i-np 13660 24887 13299 53.44 97.36 69.00 o 6349 4110 4022 97.86 63.35 76.91 b-vp 4768 157 138 87.90 2.89 5.60 i-vp 2602 32 32 100.00 1.23 2.43 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 0 0 0.00 0.00 0.00 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 0 0 0.00 0.00 0.00 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 24.08 16.19 19.36 Avg2. 46451 46451 30147 64.90 64.90 64.90 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 10555 4524 42.86 37.02 39.73 pp 4633 6710 4137 61.65 89.29 72.94 vp 4768 157 123 78.34 2.58 4.99 sbar 503 0 0 0.00 0.00 0.00 adjp 384 0 0 0.00 0.00 0.00 advp 822 0 0 0.00 0.00 0.00 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 18.29 12.89 15.12 Avg2. 23486 17422 8784 50.42 37.40 42.95 Current max chunk-based F1: 67.40 (iteration 1) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 3 Log-likelihood = -3198539.684567 Norm (log-likelihood gradient vector) = 355817.566464 Norm (lambda vector) = 24.954468 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 5866 4068 69.35 87.80 77.49 b-np 12220 9797 8070 82.37 66.04 73.31 i-np 13660 25992 13292 51.14 97.31 67.04 o 6349 4354 4254 97.70 67.00 79.49 b-vp 4768 317 262 82.65 5.49 10.30 i-vp 2602 125 125 100.00 4.80 9.17 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 0 0 0.00 0.00 0.00 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 0 0 0.00 0.00 0.00 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 24.16 16.42 19.55 Avg2. 46451 46451 30071 64.74 64.74 64.74 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 9797 4170 42.56 34.12 37.88 pp 4633 5866 4066 69.31 87.76 77.45 vp 4768 317 259 81.70 5.43 10.19 sbar 503 0 0 0.00 0.00 0.00 adjp 384 0 0 0.00 0.00 0.00 advp 822 0 0 0.00 0.00 0.00 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 19.36 12.73 15.36 Avg2. 23486 15980 8495 53.16 36.17 43.05 Current max chunk-based F1: 67.40 (iteration 1) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 4 Log-likelihood = -2205396.611759 Norm (log-likelihood gradient vector) = 422733.969401 Norm (lambda vector) = 30.443076 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4925 4212 85.52 90.91 88.14 b-np 12220 13038 10360 79.46 84.78 82.03 i-np 13660 17070 12875 75.42 94.25 83.79 o 6349 6031 5511 91.38 86.80 89.03 b-vp 4768 3694 3363 91.04 70.53 79.48 i-vp 2602 1693 1591 93.98 61.15 74.09 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 0 0 0.00 0.00 0.00 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 0 0 0.00 0.00 0.00 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 25.84 24.42 25.11 Avg2. 46451 46451 37912 81.62 81.62 81.62 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 13038 8594 65.92 70.33 68.05 pp 4633 4925 4209 85.46 90.85 88.07 vp 4768 3694 2820 76.34 59.14 66.65 sbar 503 0 0 0.00 0.00 0.00 adjp 384 0 0 0.00 0.00 0.00 advp 822 0 0 0.00 0.00 0.00 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 22.77 22.03 22.40 Avg2. 23486 21657 15623 72.14 66.52 69.22 Current max chunk-based F1: 69.22 (iteration 4) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 5 Log-likelihood = -1554251.235389 Norm (log-likelihood gradient vector) = 175459.591514 Norm (lambda vector) = 30.797786 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4846 4171 86.07 90.03 88.01 b-np 12220 12815 10714 83.61 87.68 85.59 i-np 13660 15975 12962 81.14 94.89 87.48 o 6349 5884 5592 95.04 88.08 91.42 b-vp 4768 4482 3944 88.00 82.72 85.28 i-vp 2602 2449 2125 86.77 81.67 84.14 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 0 0 0.00 0.00 0.00 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 0 0 0.00 0.00 0.00 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 26.03 26.25 26.14 Avg2. 46451 46451 39508 85.05 85.05 85.05 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12815 9364 73.07 76.63 74.81 pp 4633 4846 4168 86.01 89.96 87.94 vp 4768 4482 3481 77.67 73.01 75.26 sbar 503 0 0 0.00 0.00 0.00 adjp 384 0 0 0.00 0.00 0.00 advp 822 0 0 0.00 0.00 0.00 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 23.67 23.96 23.82 Avg2. 23486 22143 17013 76.83 72.44 74.57 Current max chunk-based F1: 74.57 (iteration 5) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 6 Log-likelihood = -1320011.680028 Norm (log-likelihood gradient vector) = 103503.924823 Norm (lambda vector) = 30.778964 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4992 4298 86.10 92.77 89.31 b-np 12220 13251 11384 85.91 93.16 89.39 i-np 13660 14106 12774 90.56 93.51 92.01 o 6349 6031 5671 94.03 89.32 91.62 b-vp 4768 5016 4276 85.25 89.68 87.41 i-vp 2602 3054 2369 77.57 91.05 83.77 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 1 1 100.00 0.12 0.24 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 0 0 0.00 0.00 0.00 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 30.97 27.48 29.12 Avg2. 46451 46451 40773 87.78 87.78 87.78 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 13251 10503 79.26 85.95 82.47 pp 4633 4992 4295 86.04 92.70 89.25 vp 4768 5016 3767 75.10 79.01 77.00 sbar 503 0 0 0.00 0.00 0.00 adjp 384 0 0 0.00 0.00 0.00 advp 822 1 1 100.00 0.12 0.24 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 34.04 25.78 29.34 Avg2. 23486 23260 18566 79.82 79.05 79.43 Current max chunk-based F1: 79.43 (iteration 6) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 7 Log-likelihood = -1195817.340584 Norm (log-likelihood gradient vector) = 97762.685670 Norm (lambda vector) = 31.244267 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 5043 4381 86.87 94.56 90.55 b-np 12220 12924 11460 88.67 93.78 91.15 i-np 13660 13979 12846 91.89 94.04 92.96 o 6349 6060 5736 94.65 90.34 92.45 b-vp 4768 4964 4283 86.28 89.83 88.02 i-vp 2602 3368 2452 72.80 94.24 82.14 b-adjp 384 0 0 0.00 0.00 0.00 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 86 80 93.02 9.73 17.62 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 27 26 96.30 5.17 9.81 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 35.52 28.58 31.68 Avg2. 46451 46451 41264 88.83 88.83 88.83 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12924 10713 82.89 87.67 85.21 pp 4633 5043 4377 86.79 94.47 90.47 vp 4768 4964 3731 75.16 78.25 76.67 sbar 503 27 26 96.30 5.17 9.81 adjp 384 0 0 0.00 0.00 0.00 advp 822 86 80 93.02 9.73 17.62 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 43.42 27.53 33.69 Avg2. 23486 23044 18927 82.13 80.59 81.35 Current max chunk-based F1: 81.35 (iteration 7) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 8 Log-likelihood = -1066946.416248 Norm (log-likelihood gradient vector) = 76639.313658 Norm (lambda vector) = 32.263182 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4829 4451 92.17 96.07 94.08 b-np 12220 12294 11493 93.48 94.05 93.77 i-np 13660 14218 13182 92.71 96.50 94.57 o 6349 5980 5790 96.82 91.20 93.92 b-vp 4768 4901 4417 90.12 92.64 91.36 i-vp 2602 3365 2485 73.85 95.50 83.29 b-adjp 384 12 12 100.00 3.12 6.06 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 568 418 73.59 50.85 60.14 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 284 266 93.66 52.88 67.60 i-adjp 152 0 0 0.00 0.00 0.00 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 40.32 33.64 36.68 Avg2. 46451 46451 42514 91.52 91.52 91.52 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12294 10917 88.80 89.34 89.07 pp 4633 4829 4445 92.05 95.94 93.95 vp 4768 4901 3925 80.09 82.32 81.19 sbar 503 284 266 93.66 52.88 67.60 adjp 384 12 12 100.00 3.12 6.06 advp 822 568 402 70.77 48.91 57.84 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 52.54 37.25 43.59 Avg2. 23486 22888 19967 87.24 85.02 86.11 Current max chunk-based F1: 86.11 (iteration 8) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 9 Log-likelihood = -767052.648036 Norm (log-likelihood gradient vector) = 47520.082932 Norm (lambda vector) = 37.628605 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4813 4476 93.00 96.61 94.77 b-np 12220 12571 11742 93.41 96.09 94.73 i-np 13660 13423 12891 96.04 94.37 95.20 o 6349 6443 6080 94.37 95.76 95.06 b-vp 4768 4893 4517 92.32 94.74 93.51 i-vp 2602 3009 2479 82.39 95.27 88.36 b-adjp 384 173 134 77.46 34.90 48.11 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 774 553 71.45 67.27 69.30 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 342 308 90.06 61.23 72.90 i-adjp 152 10 7 70.00 4.61 8.64 b-prt 126 0 0 0.00 0.00 0.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 43.02 37.04 39.81 Avg2. 46451 46451 43187 92.97 92.97 92.97 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12571 11255 89.53 92.10 90.80 pp 4633 4813 4467 92.81 96.42 94.58 vp 4768 4893 4190 85.63 87.88 86.74 sbar 503 342 308 90.06 61.23 72.90 adjp 384 173 131 75.72 34.11 47.04 advp 822 774 531 68.60 64.60 66.54 prt 126 0 0 0.00 0.00 0.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 50.24 43.63 46.70 Avg2. 23486 23566 20882 88.61 88.91 88.76 Current max chunk-based F1: 88.76 (iteration 9) Training iteration elapsed (including evaluation time): 356 seconds Iteration: 10 Log-likelihood = -643417.144498 Norm (log-likelihood gradient vector) = 52143.816434 Norm (lambda vector) = 42.107082 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4837 4514 93.32 97.43 95.33 b-np 12220 12175 11616 95.41 95.06 95.23 i-np 13660 13988 13211 94.45 96.71 95.57 o 6349 6194 5993 96.75 94.39 95.56 b-vp 4768 4844 4532 93.56 95.05 94.30 i-vp 2602 2819 2457 87.16 94.43 90.65 b-adjp 384 304 212 69.74 55.21 61.63 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 884 612 69.23 74.45 71.75 i-advp 100 0 0 0.00 0.00 0.00 b-sbar 503 353 319 90.37 63.42 74.53 i-adjp 152 35 25 71.43 16.45 26.74 b-prt 126 18 18 100.00 14.29 25.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 48.07 39.84 43.57 Avg2. 46451 46451 43509 93.67 93.67 93.67 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12175 11132 91.43 91.10 91.26 pp 4633 4837 4500 93.03 97.13 95.04 vp 4768 4844 4287 88.50 89.91 89.20 sbar 503 353 318 90.08 63.22 74.30 adjp 384 304 206 67.76 53.65 59.88 advp 822 884 583 65.95 70.92 68.35 prt 126 18 18 100.00 14.29 25.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 59.68 48.02 53.22 Avg2. 23486 23415 21044 89.87 89.60 89.74 Current max chunk-based F1: 89.74 (iteration 10) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 11 Log-likelihood = -561712.420431 Norm (log-likelihood gradient vector) = 31818.735264 Norm (lambda vector) = 46.391811 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4830 4517 93.52 97.50 95.47 b-np 12220 12246 11705 95.58 95.79 95.68 i-np 13660 13743 13165 95.79 96.38 96.08 o 6349 6328 6088 96.21 95.89 96.05 b-vp 4768 4833 4543 94.00 95.28 94.64 i-vp 2602 2732 2448 89.60 94.08 91.79 b-adjp 384 360 241 66.94 62.76 64.78 i-pp 52 0 0 0.00 0.00 0.00 b-advp 822 882 627 71.09 76.28 73.59 i-advp 100 13 9 69.23 9.00 15.93 b-sbar 503 366 329 89.89 65.41 75.72 i-adjp 152 60 41 68.33 26.97 38.68 b-prt 126 58 57 98.28 45.24 61.96 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 51.42 43.03 46.85 Avg2. 46451 46451 43770 94.23 94.23 94.23 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12246 11279 92.10 92.30 92.20 pp 4633 4830 4504 93.25 97.22 95.19 vp 4768 4833 4352 90.05 91.28 90.66 sbar 503 366 328 89.62 65.21 75.49 adjp 384 360 227 63.06 59.11 61.02 advp 822 882 607 68.82 73.84 71.24 prt 126 58 57 98.28 45.24 61.96 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 59.52 52.42 55.74 Avg2. 23486 23575 21354 90.58 90.92 90.75 Current max chunk-based F1: 90.75 (iteration 11) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 12 Log-likelihood = -524913.310144 Norm (log-likelihood gradient vector) = 26619.329487 Norm (lambda vector) = 49.402243 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4816 4529 94.04 97.76 95.86 b-np 12220 12195 11721 96.11 95.92 96.01 i-np 13660 13847 13239 95.61 96.92 96.26 o 6349 6271 6051 96.49 95.31 95.90 b-vp 4768 4831 4560 94.39 95.64 95.01 i-vp 2602 2662 2447 91.92 94.04 92.97 b-adjp 384 379 266 70.18 69.27 69.72 i-pp 52 10 10 100.00 19.23 32.26 b-advp 822 836 633 75.72 77.01 76.36 i-advp 100 34 23 67.65 23.00 34.33 b-sbar 503 383 342 89.30 67.99 77.20 i-adjp 152 86 63 73.26 41.45 52.94 b-prt 126 101 98 97.03 77.78 86.34 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 57.08 47.57 51.89 Avg2. 46451 46451 43982 94.68 94.68 94.68 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12195 11298 92.64 92.45 92.55 pp 4633 4816 4518 93.81 97.52 95.63 vp 4768 4831 4407 91.22 92.43 91.82 sbar 503 383 341 89.03 67.79 76.98 adjp 384 379 252 66.49 65.62 66.06 advp 822 836 622 74.40 75.67 75.03 prt 126 101 98 97.03 77.78 86.34 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 60.46 56.93 58.64 Avg2. 23486 23541 21536 91.48 91.70 91.59 Current max chunk-based F1: 91.59 (iteration 12) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 13 Log-likelihood = -496945.562294 Norm (log-likelihood gradient vector) = 31537.129420 Norm (lambda vector) = 53.611156 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4770 4522 94.80 97.60 96.18 b-np 12220 11953 11563 96.74 94.62 95.67 i-np 13660 14384 13421 93.31 98.25 95.71 o 6349 6104 5938 97.28 93.53 95.37 b-vp 4768 4780 4553 95.25 95.49 95.37 i-vp 2602 2656 2460 92.62 94.54 93.57 b-adjp 384 378 271 71.69 70.57 71.13 i-pp 52 26 26 100.00 50.00 66.67 b-advp 822 758 606 79.95 73.72 76.71 i-advp 100 49 33 67.35 33.00 44.30 b-sbar 503 389 348 89.46 69.18 78.03 i-adjp 152 89 69 77.53 45.39 57.26 b-prt 126 115 112 97.39 88.89 92.95 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 57.67 50.24 53.70 Avg2. 46451 46451 43922 94.56 94.56 94.56 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 11953 11074 92.65 90.62 91.62 pp 4633 4770 4519 94.74 97.54 96.12 vp 4768 4780 4415 92.36 92.60 92.48 sbar 503 389 347 89.20 68.99 77.80 adjp 384 378 259 68.52 67.45 67.98 advp 822 758 599 79.02 72.87 75.82 prt 126 115 112 97.39 88.89 92.95 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 61.39 57.90 59.59 Avg2. 23486 23143 21325 92.14 90.80 91.47 Current max chunk-based F1: 91.59 (iteration 12) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 14 Log-likelihood = -489566.140921 Norm (log-likelihood gradient vector) = 67328.904497 Norm (lambda vector) = 57.181886 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4784 4530 94.69 97.78 96.21 b-np 12220 12187 11739 96.32 96.06 96.19 i-np 13660 13931 13313 95.56 97.46 96.50 o 6349 6281 6070 96.64 95.61 96.12 b-vp 4768 4795 4565 95.20 95.74 95.47 i-vp 2602 2674 2470 92.37 94.93 93.63 b-adjp 384 377 272 72.15 70.83 71.48 i-pp 52 27 27 100.00 51.92 68.35 b-advp 822 754 605 80.24 73.60 76.78 i-advp 100 48 33 68.75 33.00 44.59 b-sbar 503 386 346 89.64 68.79 77.84 i-adjp 152 91 71 78.02 46.71 58.44 b-prt 126 116 113 97.41 89.68 93.39 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 57.85 50.61 53.99 Avg2. 46451 46451 44154 95.06 95.06 95.06 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12187 11362 93.23 92.98 93.10 pp 4633 4784 4528 94.65 97.73 96.17 vp 4768 4795 4429 92.37 92.89 92.63 sbar 503 386 345 89.38 68.59 77.62 adjp 384 377 261 69.23 67.97 68.59 advp 822 754 599 79.44 72.87 76.02 prt 126 116 113 97.41 89.68 93.39 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 61.57 58.27 59.88 Avg2. 23486 23399 21637 92.47 92.13 92.30 Current max chunk-based F1: 92.30 (iteration 14) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 15 Log-likelihood = -467994.358735 Norm (log-likelihood gradient vector) = 28947.175958 Norm (lambda vector) = 57.337733 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4783 4532 94.75 97.82 96.26 b-np 12220 12276 11798 96.11 96.55 96.33 i-np 13660 13775 13264 96.29 97.10 96.69 o 6349 6349 6113 96.28 96.28 96.28 b-vp 4768 4789 4558 95.18 95.60 95.39 i-vp 2602 2701 2480 91.82 95.31 93.53 b-adjp 384 362 269 74.31 70.05 72.12 i-pp 52 29 28 96.55 53.85 69.14 b-advp 822 750 609 81.20 74.09 77.48 i-advp 100 51 35 68.63 35.00 46.36 b-sbar 503 380 346 91.05 68.79 78.37 i-adjp 152 89 71 79.78 46.71 58.92 b-prt 126 117 113 96.58 89.68 93.00 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 57.93 50.84 54.15 Avg2. 46451 46451 44216 95.19 95.19 95.19 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12276 11446 93.24 93.67 93.45 pp 4633 4783 4531 94.73 97.80 96.24 vp 4768 4789 4420 92.29 92.70 92.50 sbar 503 380 346 91.05 68.79 78.37 adjp 384 362 259 71.55 67.45 69.44 advp 822 750 602 80.27 73.24 76.59 prt 126 117 113 96.58 89.68 93.00 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 61.97 58.33 60.10 Avg2. 23486 23457 21717 92.58 92.47 92.52 Current max chunk-based F1: 92.52 (iteration 15) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 16 Log-likelihood = -453498.541278 Norm (log-likelihood gradient vector) = 25492.686558 Norm (lambda vector) = 57.604750 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4778 4528 94.77 97.73 96.23 b-np 12220 12350 11838 95.85 96.87 96.36 i-np 13660 13663 13206 96.66 96.68 96.67 o 6349 6400 6137 95.89 96.66 96.27 b-vp 4768 4775 4552 95.33 95.47 95.40 i-vp 2602 2741 2498 91.13 96.00 93.51 b-adjp 384 346 267 77.17 69.53 73.15 i-pp 52 32 30 93.75 57.69 71.43 b-advp 822 739 607 82.14 73.84 77.77 i-advp 100 50 36 72.00 36.00 48.00 b-sbar 503 377 343 90.98 68.19 77.95 i-adjp 152 82 68 82.93 44.74 58.12 b-prt 126 118 114 96.61 90.48 93.44 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 58.26 50.99 54.39 Avg2. 46451 46451 44224 95.21 95.21 95.21 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12350 11497 93.09 94.08 93.59 pp 4633 4778 4528 94.77 97.73 96.23 vp 4768 4775 4411 92.38 92.51 92.44 sbar 503 377 343 90.98 68.19 77.95 adjp 384 346 256 73.99 66.67 70.14 advp 822 739 603 81.60 73.36 77.26 prt 126 118 114 96.61 90.48 93.44 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 62.34 58.30 60.25 Avg2. 23486 23483 21752 92.63 92.62 92.62 Current max chunk-based F1: 92.62 (iteration 16) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 17 Log-likelihood = -439830.845452 Norm (log-likelihood gradient vector) = 30936.473511 Norm (lambda vector) = 58.510479 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4771 4533 95.01 97.84 96.41 b-np 12220 12370 11855 95.84 97.01 96.42 i-np 13660 13617 13176 96.76 96.46 96.61 o 6349 6399 6137 95.91 96.66 96.28 b-vp 4768 4759 4550 95.61 95.43 95.52 i-vp 2602 2756 2501 90.75 96.12 93.36 b-adjp 384 335 268 80.00 69.79 74.55 i-pp 52 37 33 89.19 63.46 74.16 b-advp 822 762 628 82.41 76.40 79.29 i-advp 100 57 41 71.93 41.00 52.23 b-sbar 503 379 345 91.03 68.59 78.23 i-adjp 152 89 77 86.52 50.66 63.90 b-prt 126 120 116 96.67 92.06 94.31 i-sbar 12 0 0 0.00 0.00 0.00 b-conjp 16 0 0 0.00 0.00 0.00 i-conjp 24 0 0 0.00 0.00 0.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 58.38 52.07 55.05 Avg2. 46451 46451 44260 95.28 95.28 95.28 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12370 11521 93.14 94.28 93.70 pp 4633 4771 4533 95.01 97.84 96.41 vp 4768 4759 4404 92.54 92.37 92.45 sbar 503 379 345 91.03 68.59 78.23 adjp 384 335 261 77.91 67.97 72.60 advp 822 762 623 81.76 75.79 78.66 prt 126 120 116 96.67 92.06 94.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 0 0 0.00 0.00 0.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 62.81 58.89 60.78 Avg2. 23486 23496 21803 92.79 92.83 92.81 Current max chunk-based F1: 92.81 (iteration 17) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 18 Log-likelihood = -416104.118101 Norm (log-likelihood gradient vector) = 29513.779705 Norm (lambda vector) = 60.792468 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4734 4529 95.67 97.76 96.70 b-np 12220 12168 11769 96.72 96.31 96.51 i-np 13660 13973 13339 95.46 97.65 96.54 o 6349 6245 6060 97.04 95.45 96.24 b-vp 4768 4715 4552 96.54 95.47 96.00 i-vp 2602 2699 2487 92.15 95.58 93.83 b-adjp 384 340 274 80.59 71.35 75.69 i-pp 52 49 33 67.35 63.46 65.35 b-advp 822 826 654 79.18 79.56 79.37 i-advp 100 62 46 74.19 46.00 56.79 b-sbar 503 404 371 91.83 73.76 81.81 i-adjp 152 102 90 88.24 59.21 70.87 b-prt 126 126 120 95.24 95.24 95.24 i-sbar 12 2 1 50.00 8.33 14.29 b-conjp 16 2 2 100.00 12.50 22.22 i-conjp 24 4 4 100.00 16.67 28.57 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.01 55.21 61.74 Avg2. 46451 46451 44331 95.44 95.44 95.44 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12168 11393 93.63 93.23 93.43 pp 4633 4734 4528 95.65 97.73 96.68 vp 4768 4715 4417 93.68 92.64 93.16 sbar 503 404 371 91.83 73.76 81.81 adjp 384 340 269 79.12 70.05 74.31 advp 822 826 646 78.21 78.59 78.40 prt 126 126 120 95.24 95.24 95.24 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 2 2 100.00 12.50 22.22 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.74 61.37 66.57 Avg2. 23486 23315 21746 93.27 92.59 92.93 Current max chunk-based F1: 92.93 (iteration 18) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 19 Log-likelihood = -391825.475859 Norm (log-likelihood gradient vector) = 41476.254798 Norm (lambda vector) = 67.302576 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4740 4546 95.91 98.12 97.00 b-np 12220 12311 11861 96.34 97.06 96.70 i-np 13660 13637 13202 96.81 96.65 96.73 o 6349 6371 6129 96.20 96.53 96.37 b-vp 4768 4742 4576 96.50 95.97 96.24 i-vp 2602 2683 2494 92.96 95.85 94.38 b-adjp 384 343 277 80.76 72.14 76.20 i-pp 52 43 33 76.74 63.46 69.47 b-advp 822 851 668 78.50 81.27 79.86 i-advp 100 64 47 73.44 47.00 57.32 b-sbar 503 417 381 91.37 75.75 82.83 i-adjp 152 107 93 86.92 61.18 71.81 b-prt 126 126 120 95.24 95.24 95.24 i-sbar 12 2 1 50.00 8.33 14.29 b-conjp 16 5 5 100.00 31.25 47.62 i-conjp 24 9 9 100.00 37.50 54.55 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.38 57.67 63.39 Avg2. 46451 46451 44442 95.68 95.68 95.68 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12311 11513 93.52 94.21 93.86 pp 4633 4740 4546 95.91 98.12 97.00 vp 4768 4742 4457 93.99 93.48 93.73 sbar 503 417 381 91.37 75.75 82.83 adjp 384 343 273 79.59 71.09 75.10 advp 822 851 660 77.56 80.29 78.90 prt 126 126 120 95.24 95.24 95.24 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 5 4 80.00 25.00 38.10 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.72 63.32 66.81 Avg2. 23486 23535 21954 93.28 93.48 93.38 Current max chunk-based F1: 93.38 (iteration 19) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 20 Log-likelihood = -362230.478224 Norm (log-likelihood gradient vector) = 24644.409694 Norm (lambda vector) = 68.571365 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4748 4551 95.85 98.23 97.03 b-np 12220 12268 11845 96.55 96.93 96.74 i-np 13660 13692 13243 96.72 96.95 96.83 o 6349 6341 6119 96.50 96.38 96.44 b-vp 4768 4755 4587 96.47 96.20 96.34 i-vp 2602 2671 2492 93.30 95.77 94.52 b-adjp 384 349 284 81.38 73.96 77.49 i-pp 52 41 33 80.49 63.46 70.97 b-advp 822 850 670 78.82 81.51 80.14 i-advp 100 61 45 73.77 45.00 55.90 b-sbar 503 415 380 91.57 75.55 82.79 i-adjp 152 118 102 86.44 67.11 75.56 b-prt 126 126 120 95.24 95.24 95.24 i-sbar 12 2 1 50.00 8.33 14.29 b-conjp 16 5 5 100.00 31.25 47.62 i-conjp 24 9 9 100.00 37.50 54.55 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.65 57.97 63.69 Avg2. 46451 46451 44486 95.77 95.77 95.77 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12268 11502 93.76 94.12 93.94 pp 4633 4748 4551 95.85 98.23 97.03 vp 4768 4755 4475 94.11 93.85 93.98 sbar 503 415 380 91.57 75.55 82.79 adjp 384 349 281 80.52 73.18 76.67 advp 822 850 662 77.88 80.54 79.19 prt 126 126 120 95.24 95.24 95.24 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 5 4 80.00 25.00 38.10 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.89 63.57 67.03 Avg2. 23486 23516 21975 93.45 93.57 93.51 Current max chunk-based F1: 93.51 (iteration 20) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 21 Log-likelihood = -348740.139519 Norm (log-likelihood gradient vector) = 14034.897160 Norm (lambda vector) = 69.291344 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4749 4559 96.00 98.40 97.19 b-np 12220 12171 11810 97.03 96.64 96.84 i-np 13660 13859 13330 96.18 97.58 96.88 o 6349 6276 6095 97.12 96.00 96.55 b-vp 4768 4764 4595 96.45 96.37 96.41 i-vp 2602 2633 2485 94.38 95.50 94.94 b-adjp 384 361 295 81.72 76.82 79.19 i-pp 52 35 33 94.29 63.46 75.86 b-advp 822 852 678 79.58 82.48 81.00 i-advp 100 59 45 76.27 45.00 56.60 b-sbar 503 421 388 92.16 77.14 83.98 i-adjp 152 122 106 86.89 69.74 77.37 b-prt 126 126 120 95.24 95.24 95.24 i-sbar 12 2 1 50.00 8.33 14.29 b-conjp 16 7 7 100.00 43.75 60.87 i-conjp 24 14 14 100.00 58.33 73.68 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.67 60.04 65.34 Avg2. 46451 46451 44561 95.93 95.93 95.93 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12171 11468 94.22 93.85 94.03 pp 4633 4749 4559 96.00 98.40 97.19 vp 4768 4764 4491 94.27 94.19 94.23 sbar 503 421 388 92.16 77.14 83.98 adjp 384 361 292 80.89 76.04 78.39 advp 822 852 668 78.40 81.27 79.81 prt 126 126 120 95.24 95.24 95.24 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 7 100.00 43.75 60.87 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.12 65.99 69.37 Avg2. 23486 23451 21993 93.78 93.64 93.71 Current max chunk-based F1: 93.71 (iteration 21) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 22 Log-likelihood = -337249.837963 Norm (log-likelihood gradient vector) = 13877.539228 Norm (lambda vector) = 70.812409 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4736 4556 96.20 98.34 97.26 b-np 12220 12199 11832 96.99 96.82 96.91 i-np 13660 13789 13312 96.54 97.45 96.99 o 6349 6308 6119 97.00 96.38 96.69 b-vp 4768 4753 4591 96.59 96.29 96.44 i-vp 2602 2652 2493 94.00 95.81 94.90 b-adjp 384 368 301 81.79 78.39 80.05 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 835 674 80.72 82.00 81.35 i-advp 100 56 43 76.79 43.00 55.13 b-sbar 503 440 404 91.82 80.32 85.68 i-adjp 152 127 108 85.04 71.05 77.42 b-prt 126 126 120 95.24 95.24 95.24 i-sbar 12 2 1 50.00 8.33 14.29 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.94 60.98 65.58 Avg2. 46451 46451 44611 96.04 96.04 96.04 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12199 11505 94.31 94.15 94.23 pp 4633 4736 4556 96.20 98.34 97.26 vp 4768 4753 4486 94.38 94.09 94.23 sbar 503 440 404 91.82 80.32 85.68 adjp 384 368 296 80.43 77.08 78.72 advp 822 835 664 79.52 80.78 80.14 prt 126 126 120 95.24 95.24 95.24 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.08 67.00 69.45 Avg2. 23486 23466 22039 93.92 93.84 93.88 Current max chunk-based F1: 93.88 (iteration 22) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 23 Log-likelihood = -328677.008854 Norm (log-likelihood gradient vector) = 13400.640538 Norm (lambda vector) = 72.852967 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4692 4545 96.87 98.10 97.48 b-np 12220 11972 11688 97.63 95.65 96.63 i-np 13660 14267 13461 94.35 98.54 96.40 o 6349 6100 5969 97.85 94.01 95.90 b-vp 4768 4738 4584 96.75 96.14 96.44 i-vp 2602 2666 2503 93.89 96.20 95.03 b-adjp 384 372 309 83.06 80.47 81.75 i-pp 52 32 32 100.00 61.54 76.19 b-advp 822 799 670 83.85 81.51 82.67 i-advp 100 58 49 84.48 49.00 62.03 b-sbar 503 463 425 91.79 84.49 87.99 i-adjp 152 128 112 87.50 73.68 80.00 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 4 2 50.00 16.67 25.00 b-conjp 16 11 9 81.82 56.25 66.67 i-conjp 24 20 18 90.00 75.00 81.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.18 62.66 66.65 Avg2. 46451 46451 44497 95.79 95.79 95.79 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 11972 11289 94.30 92.38 93.33 pp 4633 4692 4545 96.87 98.10 97.48 vp 4768 4738 4473 94.41 93.81 94.11 sbar 503 463 423 91.36 84.10 87.58 adjp 384 372 304 81.72 79.17 80.42 advp 822 799 660 82.60 80.29 81.43 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 9 81.82 56.25 66.67 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.69 68.01 69.80 Avg2. 23486 23176 21824 94.17 92.92 93.54 Current max chunk-based F1: 93.88 (iteration 22) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 24 Log-likelihood = -315322.675408 Norm (log-likelihood gradient vector) = 32781.323098 Norm (lambda vector) = 77.454903 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4703 4549 96.73 98.19 97.45 b-np 12220 12202 11856 97.16 97.02 97.09 i-np 13660 13814 13347 96.62 97.71 97.16 o 6349 6301 6116 97.06 96.33 96.70 b-vp 4768 4742 4589 96.77 96.25 96.51 i-vp 2602 2665 2501 93.85 96.12 94.97 b-adjp 384 379 310 81.79 80.73 81.26 i-pp 52 32 31 96.88 59.62 73.81 b-advp 822 788 665 84.39 80.90 82.61 i-advp 100 58 47 81.03 47.00 59.49 b-sbar 503 463 422 91.14 83.90 87.37 i-adjp 152 139 113 81.29 74.34 77.66 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 5 2 40.00 16.67 23.53 b-conjp 16 11 9 81.82 56.25 66.67 i-conjp 24 20 18 90.00 75.00 81.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.02 62.60 66.10 Avg2. 46451 46451 44696 96.22 96.22 96.22 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12202 11539 94.57 94.43 94.50 pp 4633 4703 4549 96.73 98.19 97.45 vp 4768 4742 4476 94.39 93.88 94.13 sbar 503 463 420 90.71 83.50 86.96 adjp 384 379 304 80.21 79.17 79.69 advp 822 788 655 83.12 79.68 81.37 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 9 81.82 56.25 66.67 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.53 68.11 69.78 Avg2. 23486 23417 22073 94.26 93.98 94.12 Current max chunk-based F1: 94.12 (iteration 24) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 25 Log-likelihood = -304854.342222 Norm (log-likelihood gradient vector) = 15158.107678 Norm (lambda vector) = 80.258717 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4704 4555 96.83 98.32 97.57 b-np 12220 12221 11868 97.11 97.12 97.12 i-np 13660 13820 13360 96.67 97.80 97.23 o 6349 6304 6124 97.14 96.46 96.80 b-vp 4768 4742 4591 96.82 96.29 96.55 i-vp 2602 2671 2506 93.82 96.31 95.05 b-adjp 384 358 300 83.80 78.12 80.86 i-pp 52 34 32 94.12 61.54 74.42 b-advp 822 784 664 84.69 80.78 82.69 i-advp 100 59 48 81.36 48.00 60.38 b-sbar 503 457 420 91.90 83.50 87.50 i-adjp 152 132 112 84.85 73.68 78.87 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 5 2 40.00 16.67 23.53 b-conjp 16 11 9 81.82 56.25 66.67 i-conjp 24 20 18 90.00 75.00 81.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.24 62.59 66.20 Avg2. 46451 46451 44730 96.30 96.30 96.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12221 11563 94.62 94.62 94.62 pp 4633 4704 4555 96.83 98.32 97.57 vp 4768 4742 4482 94.52 94.00 94.26 sbar 503 457 418 91.47 83.10 87.08 adjp 384 358 294 82.12 76.56 79.25 advp 822 784 654 83.42 79.56 81.44 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 9 81.82 56.25 66.67 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.86 67.84 69.79 Avg2. 23486 23406 22096 94.40 94.08 94.24 Current max chunk-based F1: 94.24 (iteration 25) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 26 Log-likelihood = -300748.903234 Norm (log-likelihood gradient vector) = 12176.670722 Norm (lambda vector) = 79.941917 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4548 97.20 98.17 97.68 b-np 12220 12282 11907 96.95 97.44 97.19 i-np 13660 13726 13318 97.03 97.50 97.26 o 6349 6360 6150 96.70 96.87 96.78 b-vp 4768 4746 4590 96.71 96.27 96.49 i-vp 2602 2666 2502 93.85 96.16 94.99 b-adjp 384 342 295 86.26 76.82 81.27 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 799 673 84.23 81.87 83.04 i-advp 100 60 49 81.67 49.00 61.25 b-sbar 503 466 429 92.06 85.29 88.54 i-adjp 152 124 109 87.90 71.71 78.99 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 5 2 40.00 16.67 23.53 b-conjp 16 12 9 75.00 56.25 64.29 i-conjp 24 21 18 85.71 75.00 80.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.11 62.72 66.21 Avg2. 46451 46451 44753 96.34 96.34 96.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12282 11615 94.57 95.05 94.81 pp 4633 4679 4548 97.20 98.17 97.68 vp 4768 4746 4481 94.42 93.98 94.20 sbar 503 466 427 91.63 84.89 88.13 adjp 384 342 289 84.50 75.26 79.61 advp 822 799 663 82.98 80.66 81.80 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 9 75.00 56.25 64.29 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.41 68.03 69.68 Avg2. 23486 23455 22153 94.45 94.32 94.39 Current max chunk-based F1: 94.39 (iteration 26) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 27 Log-likelihood = -290503.535416 Norm (log-likelihood gradient vector) = 9667.698092 Norm (lambda vector) = 81.528561 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4683 4553 97.22 98.27 97.75 b-np 12220 12175 11846 97.30 96.94 97.12 i-np 13660 13966 13414 96.05 98.20 97.11 o 6349 6240 6084 97.50 95.83 96.66 b-vp 4768 4746 4596 96.84 96.39 96.62 i-vp 2602 2647 2501 94.48 96.12 95.29 b-adjp 384 327 288 88.07 75.00 81.01 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 826 686 83.05 83.45 83.25 i-advp 100 65 52 80.00 52.00 63.03 b-sbar 503 459 424 92.37 84.29 88.15 i-adjp 152 116 103 88.79 67.76 76.87 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 5 2 40.00 16.67 23.53 b-conjp 16 12 9 75.00 56.25 64.29 i-conjp 24 21 18 85.71 75.00 80.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.16 62.58 66.16 Avg2. 46451 46451 44730 96.30 96.30 96.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12175 11534 94.74 94.39 94.56 pp 4633 4683 4553 97.22 98.27 97.75 vp 4768 4746 4495 94.71 94.27 94.49 sbar 503 459 423 92.16 84.10 87.94 adjp 384 327 283 86.54 73.70 79.61 advp 822 826 674 81.60 82.00 81.80 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 9 75.00 56.25 64.29 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.58 67.90 69.69 Avg2. 23486 23357 22092 94.58 94.06 94.32 Current max chunk-based F1: 94.39 (iteration 26) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 28 Log-likelihood = -281126.817239 Norm (log-likelihood gradient vector) = 12168.381133 Norm (lambda vector) = 84.151306 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4691 4558 97.16 98.38 97.77 b-np 12220 12235 11888 97.16 97.28 97.22 i-np 13660 13806 13358 96.76 97.79 97.27 o 6349 6311 6128 97.10 96.52 96.81 b-vp 4768 4747 4606 97.03 96.60 96.82 i-vp 2602 2639 2500 94.73 96.08 95.40 b-adjp 384 330 290 87.88 75.52 81.23 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 852 696 81.69 84.67 83.15 i-advp 100 71 55 77.46 55.00 64.33 b-sbar 503 459 428 93.25 85.09 88.98 i-adjp 152 110 101 91.82 66.45 77.10 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 4 2 50.00 16.67 25.00 b-conjp 16 12 9 75.00 56.25 64.29 i-conjp 24 21 18 85.71 75.00 80.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.68 62.84 66.53 Avg2. 46451 46451 44791 96.43 96.43 96.43 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12235 11595 94.77 94.89 94.83 pp 4633 4691 4558 97.16 98.38 97.77 vp 4768 4747 4505 94.90 94.48 94.69 sbar 503 459 427 93.03 84.89 88.77 adjp 384 330 285 86.36 74.22 79.83 advp 822 852 684 80.28 83.21 81.72 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 9 75.00 56.25 64.29 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.53 68.24 69.84 Avg2. 23486 23455 22184 94.58 94.46 94.52 Current max chunk-based F1: 94.52 (iteration 28) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 29 Log-likelihood = -273471.663247 Norm (log-likelihood gradient vector) = 10938.898266 Norm (lambda vector) = 87.730221 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4691 4561 97.23 98.45 97.83 b-np 12220 12258 11912 97.18 97.48 97.33 i-np 13660 13691 13311 97.22 97.45 97.33 o 6349 6340 6148 96.97 96.83 96.90 b-vp 4768 4756 4611 96.95 96.71 96.83 i-vp 2602 2633 2499 94.91 96.04 95.47 b-adjp 384 341 296 86.80 77.08 81.66 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 878 708 80.64 86.13 83.29 i-advp 100 76 56 73.68 56.00 63.64 b-sbar 503 476 440 92.44 87.48 89.89 i-adjp 152 111 102 91.89 67.11 77.57 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 11 9 81.82 56.25 66.67 i-conjp 24 20 18 90.00 75.00 81.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.93 63.62 67.08 Avg2. 46451 46451 44828 96.51 96.51 96.51 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12258 11625 94.84 95.13 94.98 pp 4633 4691 4561 97.23 98.45 97.83 vp 4768 4756 4512 94.87 94.63 94.75 sbar 503 476 439 92.23 87.28 89.68 adjp 384 341 292 85.63 76.04 80.55 advp 822 878 694 79.04 84.43 81.65 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 9 81.82 56.25 66.67 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.95 68.82 70.35 Avg2. 23486 23540 22253 94.53 94.75 94.64 Current max chunk-based F1: 94.64 (iteration 29) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 30 Log-likelihood = -268133.639263 Norm (log-likelihood gradient vector) = 12488.525385 Norm (lambda vector) = 91.240749 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4704 4566 97.07 98.55 97.80 b-np 12220 12155 11864 97.61 97.09 97.35 i-np 13660 13848 13380 96.62 97.95 97.28 o 6349 6257 6098 97.46 96.05 96.75 b-vp 4768 4752 4611 97.03 96.71 96.87 i-vp 2602 2631 2500 95.02 96.08 95.55 b-adjp 384 361 307 85.04 79.95 82.42 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 889 718 80.76 87.35 83.93 i-advp 100 74 56 75.68 56.00 64.37 b-sbar 503 471 434 92.14 86.28 89.12 i-adjp 152 114 98 85.96 64.47 73.68 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.21 62.88 66.79 Avg2. 46451 46451 44813 96.47 96.47 96.47 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12155 11561 95.11 94.61 94.86 pp 4633 4704 4566 97.07 98.55 97.80 vp 4768 4752 4515 95.01 94.69 94.85 sbar 503 471 433 91.93 86.08 88.91 adjp 384 361 299 82.83 77.86 80.27 advp 822 889 701 78.85 85.28 81.94 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.35 68.31 70.27 Avg2. 23486 23470 22204 94.61 94.54 94.57 Current max chunk-based F1: 94.64 (iteration 29) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 31 Log-likelihood = -262125.518874 Norm (log-likelihood gradient vector) = 11709.352535 Norm (lambda vector) = 94.330860 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4708 4571 97.09 98.66 97.87 b-np 12220 12186 11897 97.63 97.36 97.49 i-np 13660 13770 13372 97.11 97.89 97.50 o 6349 6292 6130 97.43 96.55 96.99 b-vp 4768 4753 4615 97.10 96.79 96.94 i-vp 2602 2641 2503 94.77 96.20 95.48 b-adjp 384 368 312 84.78 81.25 82.98 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 878 716 81.55 87.10 84.24 i-advp 100 74 56 75.68 56.00 64.37 b-sbar 503 472 438 92.80 87.08 89.85 i-adjp 152 117 101 86.32 66.45 75.09 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 8 7 87.50 43.75 58.33 i-conjp 24 15 14 93.33 58.33 71.79 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.20 62.40 66.51 Avg2. 46451 46451 44889 96.64 96.64 96.64 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12186 11616 95.32 95.06 95.19 pp 4633 4708 4571 97.09 98.66 97.87 vp 4768 4753 4513 94.95 94.65 94.80 sbar 503 472 436 92.37 86.68 89.44 adjp 384 368 303 82.34 78.91 80.59 advp 822 878 698 79.50 84.91 82.12 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 8 7 87.50 43.75 58.33 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.29 67.87 70.01 Avg2. 23486 23502 22265 94.74 94.80 94.77 Current max chunk-based F1: 94.77 (iteration 31) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 32 Log-likelihood = -257074.244984 Norm (log-likelihood gradient vector) = 8575.773779 Norm (lambda vector) = 94.695088 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4699 4569 97.23 98.62 97.92 b-np 12220 12172 11906 97.81 97.43 97.62 i-np 13660 13802 13409 97.15 98.16 97.65 o 6349 6292 6138 97.55 96.68 97.11 b-vp 4768 4745 4608 97.11 96.64 96.88 i-vp 2602 2659 2508 94.32 96.39 95.34 b-adjp 384 380 313 82.37 81.51 81.94 i-pp 52 35 33 94.29 63.46 75.86 b-advp 822 844 700 82.94 85.16 84.03 i-advp 100 72 57 79.17 57.00 66.28 b-sbar 503 475 439 92.42 87.28 89.78 i-adjp 152 122 100 81.97 65.79 72.99 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 7 6 85.71 37.50 52.17 i-conjp 24 12 12 100.00 50.00 66.67 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.19 61.63 66.07 Avg2. 46451 46451 44922 96.71 96.71 96.71 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12172 11638 95.61 95.24 95.42 pp 4633 4699 4569 97.23 98.62 97.92 vp 4768 4745 4505 94.94 94.48 94.71 sbar 503 475 437 92.00 86.88 89.37 adjp 384 380 302 79.47 78.65 79.06 advp 822 844 685 81.16 83.33 82.23 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 6 85.71 37.50 52.17 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.99 67.07 69.45 Avg2. 23486 23451 22263 94.93 94.79 94.86 Current max chunk-based F1: 94.86 (iteration 32) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 33 Log-likelihood = -248818.513494 Norm (log-likelihood gradient vector) = 8733.095628 Norm (lambda vector) = 95.673449 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4708 4572 97.11 98.68 97.89 b-np 12220 12397 11970 96.56 97.95 97.25 i-np 13660 13396 13135 98.05 96.16 97.09 o 6349 6472 6194 95.70 97.56 96.62 b-vp 4768 4741 4598 96.98 96.43 96.71 i-vp 2602 2683 2514 93.70 96.62 95.14 b-adjp 384 387 314 81.14 81.77 81.45 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 816 690 84.56 83.94 84.25 i-advp 100 69 56 81.16 56.00 66.27 b-sbar 503 480 443 92.29 88.07 90.13 i-adjp 152 115 93 80.87 61.18 69.66 b-prt 126 130 121 93.08 96.03 94.53 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 5 4 80.00 25.00 38.10 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.53 59.96 64.82 Avg2. 46451 46451 44749 96.34 96.34 96.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12397 11661 94.06 95.43 94.74 pp 4633 4708 4572 97.11 98.68 97.89 vp 4768 4741 4491 94.73 94.19 94.46 sbar 503 480 441 91.88 87.67 89.73 adjp 384 387 300 77.52 78.12 77.82 advp 822 816 676 82.84 82.24 82.54 prt 126 130 121 93.08 96.03 94.53 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 5 4 80.00 25.00 38.10 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.12 65.74 68.32 Avg2. 23486 23664 22266 94.09 94.81 94.45 Current max chunk-based F1: 94.86 (iteration 32) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 34 Log-likelihood = -254291.846472 Norm (log-likelihood gradient vector) = 45726.393211 Norm (lambda vector) = 98.695367 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4705 4573 97.19 98.70 97.94 b-np 12220 12263 11948 97.43 97.77 97.60 i-np 13660 13635 13317 97.67 97.49 97.58 o 6349 6367 6167 96.86 97.13 97.00 b-vp 4768 4743 4602 97.03 96.52 96.77 i-vp 2602 2667 2509 94.08 96.43 95.24 b-adjp 384 385 314 81.56 81.77 81.66 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 835 695 83.23 84.55 83.89 i-advp 100 70 57 81.43 57.00 67.06 b-sbar 503 476 441 92.65 87.67 90.09 i-adjp 152 118 96 81.36 63.16 71.11 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 6 5 83.33 31.25 45.45 i-conjp 24 10 10 100.00 41.67 58.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.10 60.88 65.59 Avg2. 46451 46451 44892 96.64 96.64 96.64 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12263 11682 95.26 95.60 95.43 pp 4633 4705 4573 97.19 98.70 97.94 vp 4768 4743 4499 94.86 94.36 94.61 sbar 503 476 439 92.23 87.28 89.68 adjp 384 385 301 78.18 78.39 78.28 advp 822 835 681 81.56 82.85 82.20 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 6 5 83.33 31.25 45.45 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.64 66.45 68.95 Avg2. 23486 23542 22301 94.73 94.95 94.84 Current max chunk-based F1: 94.86 (iteration 32) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 35 Log-likelihood = -245652.139127 Norm (log-likelihood gradient vector) = 16793.381416 Norm (lambda vector) = 96.804890 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4697 4569 97.27 98.62 97.94 b-np 12220 12225 11934 97.62 97.66 97.64 i-np 13660 13716 13360 97.40 97.80 97.60 o 6349 6339 6157 97.13 96.98 97.05 b-vp 4768 4738 4597 97.02 96.41 96.72 i-vp 2602 2670 2508 93.93 96.39 95.14 b-adjp 384 387 313 80.88 81.51 81.19 i-pp 52 37 34 91.89 65.38 76.40 b-advp 822 825 691 83.76 84.06 83.91 i-advp 100 70 57 81.43 57.00 67.06 b-sbar 503 481 443 92.10 88.07 90.04 i-adjp 152 115 93 80.87 61.18 69.66 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 6 5 83.33 31.25 45.45 i-conjp 24 10 10 100.00 41.67 58.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.92 60.75 65.44 Avg2. 46451 46451 44895 96.65 96.65 96.65 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12225 11671 95.47 95.51 95.49 pp 4633 4697 4569 97.27 98.62 97.94 vp 4768 4738 4493 94.83 94.23 94.53 sbar 503 481 441 91.68 87.67 89.63 adjp 384 387 299 77.26 77.86 77.56 advp 822 825 677 82.06 82.36 82.21 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 6 5 83.33 31.25 45.45 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.57 66.35 68.86 Avg2. 23486 23488 22276 94.84 94.85 94.84 Current max chunk-based F1: 94.86 (iteration 32) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 36 Log-likelihood = -240223.433783 Norm (log-likelihood gradient vector) = 10248.942725 Norm (lambda vector) = 98.017397 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4692 4566 97.31 98.55 97.93 b-np 12220 12201 11922 97.71 97.56 97.64 i-np 13660 13809 13415 97.15 98.21 97.67 o 6349 6307 6146 97.45 96.80 97.12 b-vp 4768 4729 4591 97.08 96.29 96.68 i-vp 2602 2666 2507 94.04 96.35 95.18 b-adjp 384 385 312 81.04 81.25 81.14 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 813 684 84.13 83.21 83.67 i-advp 100 69 55 79.71 55.00 65.09 b-sbar 503 479 441 92.07 87.67 89.82 i-adjp 152 114 93 81.58 61.18 69.92 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 5 4 80.00 25.00 38.10 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 70.51 59.84 64.74 Avg2. 46451 46451 44902 96.67 96.67 96.67 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12201 11667 95.62 95.47 95.55 pp 4633 4692 4566 97.31 98.55 97.93 vp 4768 4729 4487 94.88 94.11 94.49 sbar 503 479 440 91.86 87.48 89.61 adjp 384 385 299 77.66 77.86 77.76 advp 822 813 669 82.29 81.39 81.83 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 5 4 80.00 25.00 38.10 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.34 65.59 68.35 Avg2. 23486 23433 22253 94.96 94.75 94.86 Current max chunk-based F1: 94.86 (iteration 32) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 37 Log-likelihood = -234107.841306 Norm (log-likelihood gradient vector) = 7513.665688 Norm (lambda vector) = 99.607709 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4681 4564 97.50 98.51 98.00 b-np 12220 12194 11919 97.74 97.54 97.64 i-np 13660 13850 13435 97.00 98.35 97.67 o 6349 6293 6141 97.58 96.72 97.15 b-vp 4768 4732 4592 97.04 96.31 96.67 i-vp 2602 2658 2503 94.17 96.20 95.17 b-adjp 384 377 308 81.70 80.21 80.95 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 821 687 83.68 83.58 83.63 i-advp 100 68 54 79.41 54.00 64.29 b-sbar 503 484 445 91.94 88.47 90.17 i-adjp 152 108 88 81.48 57.89 67.69 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 6 3 50.00 25.00 33.33 b-conjp 16 4 4 100.00 25.00 40.00 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.63 59.63 65.08 Avg2. 46451 46451 44906 96.67 96.67 96.67 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12194 11664 95.65 95.45 95.55 pp 4633 4681 4564 97.50 98.51 98.00 vp 4768 4732 4488 94.84 94.13 94.48 sbar 503 484 444 91.74 88.27 89.97 adjp 384 377 296 78.51 77.08 77.79 advp 822 821 670 81.61 81.51 81.56 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 4 4 100.00 25.00 40.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.37 65.60 69.26 Avg2. 23486 23422 22251 95.00 94.74 94.87 Current max chunk-based F1: 94.87 (iteration 37) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 38 Log-likelihood = -228454.310360 Norm (log-likelihood gradient vector) = 8379.905839 Norm (lambda vector) = 101.331627 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4678 4560 97.48 98.42 97.95 b-np 12220 12210 11931 97.71 97.64 97.67 i-np 13660 13829 13420 97.04 98.24 97.64 o 6349 6308 6147 97.45 96.82 97.13 b-vp 4768 4730 4595 97.15 96.37 96.76 i-vp 2602 2643 2495 94.40 95.89 95.14 b-adjp 384 370 304 82.16 79.17 80.64 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 840 695 82.74 84.55 83.63 i-advp 100 64 50 78.12 50.00 60.98 b-sbar 503 483 443 91.72 88.07 89.86 i-adjp 152 109 90 82.57 59.21 68.97 b-prt 126 129 121 93.80 96.03 94.90 i-sbar 12 7 4 57.14 33.33 42.11 b-conjp 16 4 4 100.00 25.00 40.00 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 71.83 59.87 65.31 Avg2. 46451 46451 44901 96.66 96.66 96.66 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12210 11676 95.63 95.55 95.59 pp 4633 4678 4560 97.48 98.42 97.95 vp 4768 4730 4489 94.90 94.15 94.53 sbar 503 483 442 91.51 87.87 89.66 adjp 384 370 294 79.46 76.56 77.98 advp 822 840 675 80.36 82.12 81.23 prt 126 129 121 93.80 96.03 94.90 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 4 4 100.00 25.00 40.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.31 65.57 69.23 Avg2. 23486 23444 22261 94.95 94.78 94.87 Current max chunk-based F1: 94.87 (iteration 37) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 39 Log-likelihood = -221589.810955 Norm (log-likelihood gradient vector) = 8876.028887 Norm (lambda vector) = 103.653397 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4683 4562 97.42 98.47 97.94 b-np 12220 12040 11800 98.01 96.56 97.28 i-np 13660 14169 13498 95.26 98.81 97.01 o 6349 6173 6042 97.88 95.16 96.50 b-vp 4768 4730 4604 97.34 96.56 96.95 i-vp 2602 2593 2472 95.33 95.00 95.17 b-adjp 384 365 303 83.01 78.91 80.91 i-pp 52 37 34 91.89 65.38 76.40 b-advp 822 866 712 82.22 86.62 84.36 i-advp 100 69 51 73.91 51.00 60.36 b-sbar 503 481 438 91.06 87.08 89.02 i-adjp 152 108 90 83.33 59.21 69.23 b-prt 126 128 121 94.53 96.03 95.28 i-sbar 12 6 4 66.67 33.33 44.44 b-conjp 16 1 1 100.00 6.25 11.76 i-conjp 24 2 2 100.00 8.33 15.38 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.39 57.64 64.18 Avg2. 46451 46451 44734 96.30 96.30 96.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12040 11455 95.14 93.74 94.44 pp 4633 4683 4562 97.42 98.47 97.94 vp 4768 4730 4507 95.29 94.53 94.90 sbar 503 481 437 90.85 86.88 88.82 adjp 384 365 294 80.55 76.56 78.50 advp 822 866 693 80.02 84.31 82.11 prt 126 128 121 94.53 96.03 95.28 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 1 1 100.00 6.25 11.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.38 63.68 68.18 Avg2. 23486 23294 22070 94.75 93.97 94.36 Current max chunk-based F1: 94.87 (iteration 37) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 40 Log-likelihood = -217713.344690 Norm (log-likelihood gradient vector) = 21053.433374 Norm (lambda vector) = 108.014907 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4691 4568 97.38 98.60 97.98 b-np 12220 12207 11929 97.72 97.62 97.67 i-np 13660 13786 13405 97.24 98.13 97.68 o 6349 6334 6162 97.28 97.05 97.17 b-vp 4768 4740 4607 97.19 96.62 96.91 i-vp 2602 2619 2487 94.96 95.58 95.27 b-adjp 384 370 306 82.70 79.69 81.17 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 857 708 82.61 86.13 84.34 i-advp 100 69 53 76.81 53.00 62.72 b-sbar 503 478 438 91.63 87.08 89.30 i-adjp 152 121 98 80.99 64.47 71.79 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 2 2 100.00 12.50 22.22 i-conjp 24 4 4 100.00 16.67 28.57 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.81 59.35 65.40 Avg2. 46451 46451 44928 96.72 96.72 96.72 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12207 11676 95.65 95.55 95.60 pp 4633 4691 4568 97.38 98.60 97.98 vp 4768 4740 4509 95.13 94.57 94.85 sbar 503 478 437 91.42 86.88 89.09 adjp 384 370 299 80.81 77.86 79.31 advp 822 857 689 80.40 83.82 82.07 prt 126 130 122 93.85 96.83 95.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 2 2 100.00 12.50 22.22 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.46 64.66 68.78 Avg2. 23486 23475 22302 95.00 94.96 94.98 Current max chunk-based F1: 94.98 (iteration 40) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 41 Log-likelihood = -210913.405023 Norm (log-likelihood gradient vector) = 7798.302111 Norm (lambda vector) = 107.380548 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4696 4570 97.32 98.64 97.97 b-np 12220 12241 11945 97.58 97.75 97.67 i-np 13660 13698 13361 97.54 97.81 97.68 o 6349 6364 6175 97.03 97.26 97.14 b-vp 4768 4740 4606 97.17 96.60 96.89 i-vp 2602 2637 2494 94.58 95.85 95.21 b-adjp 384 373 307 82.31 79.95 81.11 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 850 707 83.18 86.01 84.57 i-advp 100 67 54 80.60 54.00 64.67 b-sbar 503 474 436 91.98 86.68 89.25 i-adjp 152 127 101 79.53 66.45 72.40 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 4 4 100.00 25.00 40.00 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.06 60.96 66.46 Avg2. 46451 46451 44929 96.72 96.72 96.72 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12241 11699 95.57 95.74 95.65 pp 4633 4696 4570 97.32 98.64 97.97 vp 4768 4740 4503 95.00 94.44 94.72 sbar 503 474 435 91.77 86.48 89.05 adjp 384 373 300 80.43 78.12 79.26 advp 822 850 690 81.18 83.94 82.54 prt 126 130 122 93.85 96.83 95.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 4 4 100.00 25.00 40.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.51 65.92 69.51 Avg2. 23486 23508 22323 94.96 95.05 95.00 Current max chunk-based F1: 95.00 (iteration 41) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 42 Log-likelihood = -208569.715384 Norm (log-likelihood gradient vector) = 6008.123029 Norm (lambda vector) = 107.262320 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4704 4575 97.26 98.75 98.00 b-np 12220 12269 11957 97.46 97.85 97.65 i-np 13660 13627 13330 97.82 97.58 97.70 o 6349 6388 6190 96.90 97.50 97.20 b-vp 4768 4736 4604 97.21 96.56 96.89 i-vp 2602 2656 2506 94.35 96.31 95.32 b-adjp 384 375 308 82.13 80.21 81.16 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 838 700 83.53 85.16 84.34 i-advp 100 69 56 81.16 56.00 66.27 b-sbar 503 472 438 92.80 87.08 89.85 i-adjp 152 133 106 79.70 69.74 74.39 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 4 4 100.00 25.00 40.00 i-conjp 24 8 8 100.00 33.33 50.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.14 61.25 66.67 Avg2. 46451 46451 44943 96.75 96.75 96.75 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12269 11719 95.52 95.90 95.71 pp 4633 4704 4575 97.26 98.75 98.00 vp 4768 4736 4502 95.06 94.42 94.74 sbar 503 472 437 92.58 86.88 89.64 adjp 384 375 303 80.80 78.91 79.84 advp 822 838 687 81.98 83.58 82.77 prt 126 130 122 93.85 96.83 95.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 4 4 100.00 25.00 40.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.70 66.03 69.65 Avg2. 23486 23528 22349 94.99 95.16 95.07 Current max chunk-based F1: 95.07 (iteration 42) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 43 Log-likelihood = -205769.428388 Norm (log-likelihood gradient vector) = 8206.384811 Norm (lambda vector) = 107.931186 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4696 4572 97.36 98.68 98.02 b-np 12220 12186 11925 97.86 97.59 97.72 i-np 13660 13757 13398 97.39 98.08 97.73 o 6349 6320 6159 97.45 97.01 97.23 b-vp 4768 4742 4605 97.11 96.58 96.85 i-vp 2602 2673 2513 94.01 96.58 95.28 b-adjp 384 378 309 81.75 80.47 81.10 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 824 695 84.34 84.55 84.45 i-advp 100 74 61 82.43 61.00 70.11 b-sbar 503 479 444 92.69 88.27 90.43 i-adjp 152 137 109 79.56 71.71 75.43 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 5 5 100.00 31.25 47.62 i-conjp 24 10 10 100.00 41.67 58.82 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.25 62.27 67.32 Avg2. 46451 46451 44965 96.80 96.80 96.80 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12186 11684 95.88 95.61 95.75 pp 4633 4696 4572 97.36 98.68 98.02 vp 4768 4742 4503 94.96 94.44 94.70 sbar 503 479 443 92.48 88.07 90.22 adjp 384 378 304 80.42 79.17 79.79 advp 822 824 686 83.25 83.45 83.35 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 5 5 100.00 31.25 47.62 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.89 66.75 70.14 Avg2. 23486 23439 22319 95.22 95.03 95.13 Current max chunk-based F1: 95.13 (iteration 43) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 44 Log-likelihood = -202413.141919 Norm (log-likelihood gradient vector) = 12371.996689 Norm (lambda vector) = 109.421805 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4698 4572 97.32 98.68 98.00 b-np 12220 12229 11939 97.63 97.70 97.66 i-np 13660 13678 13355 97.64 97.77 97.70 o 6349 6351 6171 97.17 97.20 97.18 b-vp 4768 4743 4606 97.11 96.60 96.86 i-vp 2602 2681 2515 93.81 96.66 95.21 b-adjp 384 377 310 82.23 80.73 81.47 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 815 694 85.15 84.43 84.79 i-advp 100 75 62 82.67 62.00 70.86 b-sbar 503 477 442 92.66 87.87 90.20 i-adjp 152 138 110 79.71 72.37 75.86 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 8 6 75.00 50.00 60.00 b-conjp 16 6 6 100.00 37.50 54.55 i-conjp 24 12 12 100.00 50.00 66.67 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.49 63.49 68.12 Avg2. 46451 46451 44955 96.78 96.78 96.78 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12229 11696 95.64 95.71 95.68 pp 4633 4698 4572 97.32 98.68 98.00 vp 4768 4743 4502 94.92 94.42 94.67 sbar 503 477 441 92.45 87.67 90.00 adjp 384 377 305 80.90 79.43 80.16 advp 822 815 684 83.93 83.21 83.57 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 6 6 100.00 37.50 54.55 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.97 67.35 70.50 Avg2. 23486 23474 22328 95.12 95.07 95.09 Current max chunk-based F1: 95.13 (iteration 43) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 45 Log-likelihood = -198405.649723 Norm (log-likelihood gradient vector) = 7072.136290 Norm (lambda vector) = 110.852999 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4701 4573 97.28 98.70 97.99 b-np 12220 12232 11940 97.61 97.71 97.66 i-np 13660 13693 13363 97.59 97.83 97.71 o 6349 6344 6169 97.24 97.16 97.20 b-vp 4768 4745 4608 97.11 96.64 96.88 i-vp 2602 2683 2517 93.81 96.73 95.25 b-adjp 384 371 307 82.75 79.95 81.32 i-pp 52 34 33 97.06 63.46 76.74 b-advp 822 809 691 85.41 84.06 84.73 i-advp 100 75 62 82.67 62.00 70.86 b-sbar 503 474 439 92.62 87.28 89.87 i-adjp 152 134 107 79.85 70.39 74.83 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 7 6 85.71 37.50 52.17 i-conjp 24 13 12 92.31 50.00 64.86 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.25 62.90 67.25 Avg2. 46451 46451 44954 96.78 96.78 96.78 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11694 95.60 95.70 95.65 pp 4633 4701 4573 97.28 98.70 97.99 vp 4768 4745 4503 94.90 94.44 94.67 sbar 503 474 438 92.41 87.08 89.66 adjp 384 371 302 81.40 78.65 80.00 advp 822 809 681 84.18 82.85 83.51 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 6 85.71 37.50 52.17 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.61 67.17 69.78 Avg2. 23486 23468 22319 95.10 95.03 95.07 Current max chunk-based F1: 95.13 (iteration 43) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 46 Log-likelihood = -194398.806552 Norm (log-likelihood gradient vector) = 6539.011648 Norm (lambda vector) = 112.094707 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4694 4574 97.44 98.73 98.08 b-np 12220 12238 11943 97.59 97.73 97.66 i-np 13660 13697 13366 97.58 97.85 97.72 o 6349 6338 6167 97.30 97.13 97.22 b-vp 4768 4752 4616 97.14 96.81 96.97 i-vp 2602 2680 2518 93.96 96.77 95.34 b-adjp 384 365 305 83.56 79.43 81.44 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 806 693 85.98 84.31 85.14 i-advp 100 72 61 84.72 61.00 70.93 b-sbar 503 478 443 92.68 88.07 90.32 i-adjp 152 134 108 80.60 71.05 75.52 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.74 64.47 68.35 Avg2. 46451 46451 44979 96.83 96.83 96.83 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12238 11699 95.60 95.74 95.67 pp 4633 4694 4574 97.44 98.73 98.08 vp 4768 4752 4511 94.93 94.61 94.77 sbar 503 478 442 92.47 87.87 90.11 adjp 384 365 300 82.19 78.12 80.11 advp 822 806 682 84.62 82.97 83.78 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.07 68.49 70.70 Avg2. 23486 23471 22338 95.17 95.11 95.14 Current max chunk-based F1: 95.14 (iteration 46) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 47 Log-likelihood = -189824.413140 Norm (log-likelihood gradient vector) = 7753.530049 Norm (lambda vector) = 113.935272 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4569 97.65 98.62 98.13 b-np 12220 12140 11878 97.84 97.20 97.52 i-np 13660 13943 13462 96.55 98.55 97.54 o 6349 6230 6100 97.91 96.08 96.99 b-vp 4768 4747 4615 97.22 96.79 97.00 i-vp 2602 2672 2519 94.27 96.81 95.53 b-adjp 384 355 303 85.35 78.91 82.00 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 809 695 85.91 84.55 85.22 i-advp 100 74 61 82.43 61.00 70.11 b-sbar 503 474 440 92.83 87.48 90.07 i-adjp 152 130 108 83.08 71.05 76.60 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 6 4 66.67 33.33 44.44 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.63 64.15 68.13 Avg2. 46451 46451 44936 96.74 96.74 96.74 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12140 11603 95.58 94.95 95.26 pp 4633 4679 4569 97.65 98.62 98.13 vp 4768 4747 4513 95.07 94.65 94.86 sbar 503 474 439 92.62 87.28 89.87 adjp 384 355 298 83.94 77.60 80.65 advp 822 809 686 84.80 83.45 84.12 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.31 68.34 70.74 Avg2. 23486 23342 22238 95.27 94.69 94.98 Current max chunk-based F1: 95.14 (iteration 46) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 48 Log-likelihood = -188433.889656 Norm (log-likelihood gradient vector) = 20649.193495 Norm (lambda vector) = 116.938525 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4684 4572 97.61 98.68 98.14 b-np 12220 12184 11919 97.83 97.54 97.68 i-np 13660 13835 13444 97.17 98.42 97.79 o 6349 6279 6138 97.75 96.68 97.21 b-vp 4768 4749 4617 97.22 96.83 97.03 i-vp 2602 2675 2519 94.17 96.81 95.47 b-adjp 384 360 306 85.00 79.69 82.26 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 808 694 85.89 84.43 85.15 i-advp 100 74 61 82.43 61.00 70.11 b-sbar 503 476 441 92.65 87.67 90.09 i-adjp 152 131 108 82.44 71.05 76.33 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 6 4 66.67 33.33 44.44 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.58 64.05 68.05 Avg2. 46451 46451 45003 96.88 96.88 96.88 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12184 11674 95.81 95.53 95.67 pp 4633 4684 4572 97.61 98.68 98.14 vp 4768 4749 4514 95.05 94.67 94.86 sbar 503 476 440 92.44 87.48 89.89 adjp 384 360 301 83.61 78.39 80.91 advp 822 808 684 84.65 83.21 83.93 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.26 68.48 70.79 Avg2. 23486 23399 22315 95.37 95.01 95.19 Current max chunk-based F1: 95.19 (iteration 48) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 49 Log-likelihood = -186740.306059 Norm (log-likelihood gradient vector) = 10069.400776 Norm (lambda vector) = 115.368655 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4681 4570 97.63 98.64 98.13 b-np 12220 12201 11937 97.84 97.68 97.76 i-np 13660 13797 13440 97.41 98.39 97.90 o 6349 6298 6155 97.73 96.94 97.34 b-vp 4768 4750 4620 97.26 96.90 97.08 i-vp 2602 2671 2519 94.31 96.81 95.54 b-adjp 384 361 307 85.04 79.95 82.42 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 815 698 85.64 84.91 85.28 i-advp 100 74 61 82.43 61.00 70.11 b-sbar 503 476 442 92.86 87.87 90.30 i-adjp 152 131 108 82.44 71.05 76.33 b-prt 126 129 122 94.57 96.83 95.69 i-sbar 12 6 4 66.67 33.33 44.44 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.60 64.12 68.10 Avg2. 46451 46451 45041 96.96 96.96 96.96 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12201 11707 95.95 95.80 95.88 pp 4633 4681 4570 97.63 98.64 98.13 vp 4768 4750 4519 95.14 94.78 94.96 sbar 503 476 441 92.65 87.67 90.09 adjp 384 361 302 83.66 78.65 81.07 advp 822 815 687 84.29 83.58 83.93 prt 126 129 122 94.57 96.83 95.69 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.28 68.59 70.86 Avg2. 23486 23422 22356 95.45 95.19 95.32 Current max chunk-based F1: 95.32 (iteration 49) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 50 Log-likelihood = -184461.783915 Norm (log-likelihood gradient vector) = 6527.403795 Norm (lambda vector) = 116.042850 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4570 97.67 98.64 98.15 b-np 12220 12211 11946 97.83 97.76 97.79 i-np 13660 13780 13436 97.50 98.36 97.93 o 6349 6307 6164 97.73 97.09 97.41 b-vp 4768 4749 4623 97.35 96.96 97.15 i-vp 2602 2662 2516 94.52 96.69 95.59 b-adjp 384 360 311 86.39 80.99 83.60 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 824 705 85.56 85.77 85.66 i-advp 100 75 62 82.67 62.00 70.86 b-sbar 503 478 442 92.47 87.87 90.11 i-adjp 152 128 110 85.94 72.37 78.57 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 5 3 60.00 25.00 35.29 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.49 64.11 68.04 Avg2. 46451 46451 45070 97.03 97.03 97.03 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12211 11716 95.95 95.88 95.91 pp 4633 4679 4570 97.67 98.64 98.15 vp 4768 4749 4523 95.24 94.86 95.05 sbar 503 478 441 92.26 87.67 89.91 adjp 384 360 307 85.28 79.95 82.53 advp 822 824 696 84.47 84.67 84.57 prt 126 130 122 93.85 96.83 95.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.36 68.85 71.03 Avg2. 23486 23440 22383 95.49 95.30 95.40 Current max chunk-based F1: 95.40 (iteration 50) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 51 Log-likelihood = -181799.246480 Norm (log-likelihood gradient vector) = 5911.691762 Norm (lambda vector) = 116.941510 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4677 4569 97.69 98.62 98.15 b-np 12220 12229 11952 97.73 97.81 97.77 i-np 13660 13743 13412 97.59 98.18 97.89 o 6349 6329 6172 97.52 97.21 97.37 b-vp 4768 4749 4624 97.37 96.98 97.17 i-vp 2602 2661 2517 94.59 96.73 95.65 b-adjp 384 360 313 86.94 81.51 84.14 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 821 706 85.99 85.89 85.94 i-advp 100 75 63 84.00 63.00 72.00 b-sbar 503 475 440 92.63 87.48 89.98 i-adjp 152 130 110 84.62 72.37 78.01 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 6 4 66.67 33.33 44.44 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.96 65.32 68.93 Avg2. 46451 46451 45067 97.02 97.02 97.02 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12229 11720 95.84 95.91 95.87 pp 4633 4677 4569 97.69 98.62 98.15 vp 4768 4749 4526 95.30 94.92 95.11 sbar 503 475 439 92.42 87.28 89.78 adjp 384 360 308 85.56 80.21 82.80 advp 822 821 698 85.02 84.91 84.97 prt 126 130 122 93.85 96.83 95.31 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.57 69.49 71.47 Avg2. 23486 23451 22391 95.48 95.34 95.41 Current max chunk-based F1: 95.41 (iteration 51) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 52 Log-likelihood = -179922.370058 Norm (log-likelihood gradient vector) = 6428.085749 Norm (lambda vector) = 117.668993 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4565 97.69 98.53 98.11 b-np 12220 12158 11915 98.00 97.50 97.75 i-np 13660 13885 13464 96.97 98.57 97.76 o 6349 6268 6135 97.88 96.63 97.25 b-vp 4768 4742 4624 97.51 96.98 97.25 i-vp 2602 2669 2522 94.49 96.93 95.69 b-adjp 384 361 313 86.70 81.51 84.03 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 820 706 86.10 85.89 85.99 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 478 442 92.47 87.87 90.11 i-adjp 152 121 106 87.60 69.74 77.66 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.51 65.65 69.36 Avg2. 46451 46451 45046 96.98 96.98 96.98 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12158 11670 95.99 95.50 95.74 pp 4633 4673 4565 97.69 98.53 98.11 vp 4768 4742 4524 95.40 94.88 95.14 sbar 503 478 441 92.26 87.67 89.91 adjp 384 361 307 85.04 79.95 82.42 advp 822 820 698 85.12 84.91 85.02 prt 126 131 123 93.89 97.62 95.72 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.54 69.53 71.48 Avg2. 23486 23373 22337 95.57 95.11 95.34 Current max chunk-based F1: 95.41 (iteration 51) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 53 Log-likelihood = -176277.930657 Norm (log-likelihood gradient vector) = 10119.207287 Norm (lambda vector) = 119.850402 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4683 4571 97.61 98.66 98.13 b-np 12220 12263 11968 97.59 97.94 97.77 i-np 13660 13662 13362 97.80 97.82 97.81 o 6349 6361 6181 97.17 97.35 97.26 b-vp 4768 4741 4621 97.47 96.92 97.19 i-vp 2602 2674 2522 94.32 96.93 95.60 b-adjp 384 366 313 85.52 81.51 83.47 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 820 707 86.22 86.01 86.11 i-advp 100 75 65 86.67 65.00 74.29 b-sbar 503 477 441 92.45 87.67 90.00 i-adjp 152 124 105 84.68 69.08 76.09 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.11 65.73 69.23 Avg2. 46451 46451 45047 96.98 96.98 96.98 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12263 11730 95.65 95.99 95.82 pp 4633 4683 4571 97.61 98.66 98.13 vp 4768 4741 4519 95.32 94.78 95.05 sbar 503 477 440 92.24 87.48 89.80 adjp 384 366 307 83.88 79.95 81.87 advp 822 820 698 85.12 84.91 85.02 prt 126 131 123 93.89 97.62 95.72 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.37 69.56 71.42 Avg2. 23486 23491 22397 95.34 95.36 95.35 Current max chunk-based F1: 95.41 (iteration 51) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 54 Log-likelihood = -174607.898470 Norm (log-likelihood gradient vector) = 11337.570916 Norm (lambda vector) = 120.878722 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4681 4571 97.65 98.66 98.15 b-np 12220 12228 11962 97.82 97.89 97.86 i-np 13660 13736 13414 97.66 98.20 97.93 o 6349 6329 6175 97.57 97.26 97.41 b-vp 4768 4742 4622 97.47 96.94 97.20 i-vp 2602 2673 2522 94.35 96.93 95.62 b-adjp 384 367 313 85.29 81.51 83.36 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 816 705 86.40 85.77 86.08 i-advp 100 73 64 87.67 64.00 73.99 b-sbar 503 477 442 92.66 87.87 90.20 i-adjp 152 125 105 84.00 69.08 75.81 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.29 65.69 69.29 Avg2. 46451 46451 45086 97.06 97.06 97.06 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11739 96.00 96.06 96.03 pp 4633 4681 4571 97.65 98.66 98.15 vp 4768 4742 4521 95.34 94.82 95.08 sbar 503 477 441 92.45 87.67 90.00 adjp 384 367 307 83.65 79.95 81.76 advp 822 816 697 85.42 84.79 85.10 prt 126 131 123 93.89 97.62 95.72 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.44 69.58 71.46 Avg2. 23486 23452 22408 95.55 95.41 95.48 Current max chunk-based F1: 95.48 (iteration 54) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 55 Log-likelihood = -173212.139094 Norm (log-likelihood gradient vector) = 5750.642774 Norm (lambda vector) = 120.375673 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4683 4573 97.65 98.70 98.18 b-np 12220 12192 11952 98.03 97.81 97.92 i-np 13660 13810 13456 97.44 98.51 97.97 o 6349 6297 6163 97.87 97.07 97.47 b-vp 4768 4740 4621 97.49 96.92 97.20 i-vp 2602 2677 2523 94.25 96.96 95.59 b-adjp 384 367 314 85.56 81.77 83.62 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 810 703 86.79 85.52 86.15 i-advp 100 74 65 87.84 65.00 74.71 b-sbar 503 477 442 92.66 87.87 90.20 i-adjp 152 120 105 87.50 69.08 77.21 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 7 5 71.43 41.67 52.63 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.52 65.75 69.42 Avg2. 46451 46451 45108 97.11 97.11 97.11 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12192 11730 96.21 95.99 96.10 pp 4633 4683 4573 97.65 98.70 98.18 vp 4768 4740 4520 95.36 94.80 95.08 sbar 503 477 441 92.45 87.67 90.00 adjp 384 367 309 84.20 80.47 82.29 advp 822 810 696 85.93 84.67 85.29 prt 126 131 123 93.89 97.62 95.72 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.57 69.62 71.54 Avg2. 23486 23410 22401 95.69 95.38 95.53 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 56 Log-likelihood = -171673.495207 Norm (log-likelihood gradient vector) = 4037.740800 Norm (lambda vector) = 120.120710 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4682 4573 97.67 98.70 98.19 b-np 12220 12190 11949 98.02 97.78 97.90 i-np 13660 13825 13456 97.33 98.51 97.92 o 6349 6291 6160 97.92 97.02 97.47 b-vp 4768 4741 4624 97.53 96.98 97.26 i-vp 2602 2680 2526 94.25 97.08 95.65 b-adjp 384 363 311 85.67 80.99 83.27 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 802 700 87.28 85.16 86.21 i-advp 100 74 65 87.84 65.00 74.71 b-sbar 503 480 446 92.92 88.67 90.74 i-adjp 152 116 103 88.79 67.76 76.87 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.56 66.50 69.85 Avg2. 46451 46451 45106 97.10 97.10 97.10 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12190 11720 96.14 95.91 96.03 pp 4633 4682 4573 97.67 98.70 98.19 vp 4768 4741 4523 95.40 94.86 95.13 sbar 503 480 446 92.92 88.67 90.74 adjp 384 363 306 84.30 79.69 81.93 advp 822 802 693 86.41 84.31 85.34 prt 126 131 123 93.89 97.62 95.72 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.67 69.60 71.58 Avg2. 23486 23399 22393 95.70 95.35 95.52 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 57 Log-likelihood = -169580.077654 Norm (log-likelihood gradient vector) = 5620.778249 Norm (lambda vector) = 120.376499 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4682 4573 97.67 98.70 98.19 b-np 12220 12184 11935 97.96 97.67 97.81 i-np 13660 13853 13457 97.14 98.51 97.82 o 6349 6278 6150 97.96 96.87 97.41 b-vp 4768 4737 4624 97.61 96.98 97.30 i-vp 2602 2672 2525 94.50 97.04 95.75 b-adjp 384 364 311 85.44 80.99 83.16 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 796 695 87.31 84.55 85.91 i-advp 100 74 64 86.49 64.00 73.56 b-sbar 503 485 447 92.16 88.87 90.49 i-adjp 152 118 104 88.14 68.42 77.04 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.42 66.49 69.78 Avg2. 46451 46451 45079 97.05 97.05 97.05 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12184 11695 95.99 95.70 95.84 pp 4633 4682 4573 97.67 98.70 98.19 vp 4768 4737 4527 95.57 94.95 95.26 sbar 503 485 447 92.16 88.87 90.49 adjp 384 364 306 84.07 79.69 81.82 advp 822 796 687 86.31 83.58 84.92 prt 126 132 124 93.94 98.41 96.12 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.57 69.61 71.54 Avg2. 23486 23390 22368 95.63 95.24 95.43 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 58 Log-likelihood = -165812.739436 Norm (log-likelihood gradient vector) = 5947.377668 Norm (lambda vector) = 121.402475 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4562 97.73 98.47 98.10 b-np 12220 12063 11808 97.89 96.63 97.25 i-np 13660 14144 13480 95.31 98.68 96.96 o 6349 6169 6039 97.89 95.12 96.49 b-vp 4768 4730 4615 97.57 96.79 97.18 i-vp 2602 2634 2504 95.06 96.23 95.65 b-adjp 384 358 308 86.03 80.21 83.02 i-pp 52 41 36 87.80 69.23 77.42 b-advp 822 812 697 85.84 84.79 85.31 i-advp 100 68 60 88.24 60.00 71.43 b-sbar 503 484 444 91.74 88.27 89.97 i-adjp 152 115 103 89.57 67.76 77.15 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 8 7 87.50 43.75 58.33 i-conjp 24 15 14 93.33 58.33 71.79 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 72.77 64.55 68.42 Avg2. 46451 46451 44808 96.46 96.46 96.46 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12063 11476 95.13 93.91 94.52 pp 4633 4668 4562 97.73 98.47 98.10 vp 4768 4730 4514 95.43 94.67 95.05 sbar 503 484 444 91.74 88.27 89.97 adjp 384 358 302 84.36 78.65 81.40 advp 822 812 688 84.73 83.70 84.21 prt 126 132 124 93.94 98.41 96.12 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 8 7 87.50 43.75 58.33 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.06 67.98 70.43 Avg2. 23486 23255 22117 95.11 94.17 94.64 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 59 Log-likelihood = -175592.412427 Norm (log-likelihood gradient vector) = 40059.352909 Norm (lambda vector) = 126.104611 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4675 4569 97.73 98.62 98.17 b-np 12220 12141 11912 98.11 97.48 97.80 i-np 13660 13934 13483 96.76 98.70 97.72 o 6349 6248 6128 98.08 96.52 97.29 b-vp 4768 4736 4626 97.68 97.02 97.35 i-vp 2602 2665 2523 94.67 96.96 95.80 b-adjp 384 365 313 85.75 81.51 83.58 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 801 700 87.39 85.16 86.26 i-advp 100 73 64 87.67 64.00 73.99 b-sbar 503 488 448 91.80 89.07 90.41 i-adjp 152 117 104 88.89 68.42 77.32 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.53 66.53 69.86 Avg2. 46451 46451 45064 97.01 97.01 97.01 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12141 11659 96.03 95.41 95.72 pp 4633 4675 4569 97.73 98.62 98.17 vp 4768 4736 4529 95.63 94.99 95.31 sbar 503 488 448 91.80 89.07 90.41 adjp 384 365 308 84.38 80.21 82.24 advp 822 801 692 86.39 84.18 85.27 prt 126 132 124 93.94 98.41 96.12 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.59 69.71 71.60 Avg2. 23486 23348 22338 95.67 95.11 95.39 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 60 Log-likelihood = -164092.827437 Norm (log-likelihood gradient vector) = 11963.475975 Norm (lambda vector) = 122.726999 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4672 4567 97.75 98.58 98.16 b-np 12220 12195 11939 97.90 97.70 97.80 i-np 13660 13839 13454 97.22 98.49 97.85 o 6349 6293 6156 97.82 96.96 97.39 b-vp 4768 4739 4628 97.66 97.06 97.36 i-vp 2602 2662 2521 94.70 96.89 95.78 b-adjp 384 363 313 86.23 81.51 83.80 i-pp 52 37 36 97.30 69.23 80.90 b-advp 822 804 700 87.06 85.16 86.10 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 487 447 91.79 88.87 90.30 i-adjp 152 117 105 89.74 69.08 78.07 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.57 66.53 69.87 Avg2. 46451 46451 45087 97.06 97.06 97.06 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12195 11706 95.99 95.79 95.89 pp 4633 4672 4567 97.75 98.58 98.16 vp 4768 4739 4531 95.61 95.03 95.32 sbar 503 487 447 91.79 88.87 90.30 adjp 384 363 308 84.85 80.21 82.46 advp 822 804 691 85.95 84.06 84.99 prt 126 132 124 93.94 98.41 96.12 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.59 69.72 71.60 Avg2. 23486 23402 22383 95.65 95.30 95.47 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 61 Log-likelihood = -161217.628237 Norm (log-likelihood gradient vector) = 5209.776545 Norm (lambda vector) = 123.822471 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4566 97.77 98.55 98.16 b-np 12220 12217 11945 97.77 97.75 97.76 i-np 13660 13802 13429 97.30 98.31 97.80 o 6349 6309 6161 97.65 97.04 97.35 b-vp 4768 4739 4624 97.57 96.98 97.28 i-vp 2602 2650 2514 94.87 96.62 95.73 b-adjp 384 364 313 85.99 81.51 83.69 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 814 700 86.00 85.16 85.57 i-advp 100 71 63 88.73 63.00 73.68 b-sbar 503 490 447 91.22 88.87 90.03 i-adjp 152 115 103 89.57 67.76 77.15 b-prt 126 133 124 93.23 98.41 95.75 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.36 66.44 69.73 Avg2. 46451 46451 45059 97.00 97.00 97.00 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12217 11714 95.88 95.86 95.87 pp 4633 4670 4566 97.77 98.55 98.16 vp 4768 4739 4527 95.53 94.95 95.24 sbar 503 490 447 91.22 88.87 90.03 adjp 384 364 307 84.34 79.95 82.09 advp 822 814 692 85.01 84.18 84.60 prt 126 133 124 93.23 98.41 95.75 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.30 69.70 71.46 Avg2. 23486 23437 22386 95.52 95.32 95.42 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 62 Log-likelihood = -159530.856886 Norm (log-likelihood gradient vector) = 3760.058879 Norm (lambda vector) = 125.022583 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4565 97.79 98.53 98.16 b-np 12220 12223 11947 97.74 97.77 97.75 i-np 13660 13784 13421 97.37 98.25 97.81 o 6349 6321 6166 97.55 97.12 97.33 b-vp 4768 4738 4622 97.55 96.94 97.24 i-vp 2602 2647 2511 94.86 96.50 95.68 b-adjp 384 365 315 86.30 82.03 84.11 i-pp 52 36 35 97.22 67.31 79.55 b-advp 822 824 701 85.07 85.28 85.18 i-advp 100 71 61 85.92 61.00 71.35 b-sbar 503 488 446 91.39 88.67 90.01 i-adjp 152 114 103 90.35 67.76 77.44 b-prt 126 133 124 93.23 98.41 95.75 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.35 66.26 69.63 Avg2. 46451 46451 45051 96.99 96.99 96.99 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12223 11715 95.84 95.87 95.86 pp 4633 4668 4565 97.79 98.53 98.16 vp 4768 4738 4525 95.50 94.90 95.20 sbar 503 488 446 91.39 88.67 90.01 adjp 384 365 309 84.66 80.47 82.51 advp 822 824 690 83.74 83.94 83.84 prt 126 133 124 93.23 98.41 95.75 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.22 69.70 71.42 Avg2. 23486 23449 22383 95.45 95.30 95.38 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 63 Log-likelihood = -157947.487703 Norm (log-likelihood gradient vector) = 4492.073178 Norm (lambda vector) = 125.964718 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4564 97.86 98.51 98.18 b-np 12220 12234 11958 97.74 97.86 97.80 i-np 13660 13749 13405 97.50 98.13 97.81 o 6349 6335 6170 97.40 97.18 97.29 b-vp 4768 4737 4623 97.59 96.96 97.28 i-vp 2602 2651 2516 94.91 96.69 95.79 b-adjp 384 367 316 86.10 82.29 84.15 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 834 707 84.77 86.01 85.39 i-advp 100 69 60 86.96 60.00 71.01 b-sbar 503 488 449 92.01 89.26 90.62 i-adjp 152 114 104 91.23 68.42 78.20 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 0 0 0.00 0.00 0.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.56 66.65 69.94 Avg2. 46451 46451 45065 97.02 97.02 97.02 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12234 11728 95.86 95.97 95.92 pp 4633 4664 4564 97.86 98.51 98.18 vp 4768 4737 4530 95.63 95.01 95.32 sbar 503 488 449 92.01 89.26 90.62 adjp 384 367 310 84.47 80.73 82.56 advp 822 834 696 83.45 84.67 84.06 prt 126 134 124 92.54 98.41 95.38 lst 10 0 0 0.00 0.00 0.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 73.18 69.88 71.49 Avg2. 23486 23468 22410 95.49 95.42 95.46 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 64 Log-likelihood = -155564.342406 Norm (log-likelihood gradient vector) = 5397.875758 Norm (lambda vector) = 127.283975 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4677 4568 97.67 98.60 98.13 b-np 12220 12112 11875 98.04 97.18 97.61 i-np 13660 13978 13460 96.29 98.54 97.40 o 6349 6239 6097 97.72 96.03 96.87 b-vp 4768 4733 4615 97.51 96.79 97.15 i-vp 2602 2650 2509 94.68 96.43 95.54 b-adjp 384 368 317 86.14 82.55 84.31 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 832 706 84.86 85.89 85.37 i-advp 100 70 59 84.29 59.00 69.41 b-sbar 503 478 446 93.31 88.67 90.93 i-adjp 152 112 102 91.07 67.11 77.27 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 7 6 85.71 37.50 52.17 i-conjp 24 13 12 92.31 50.00 64.86 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 3 3 100.00 30.00 46.15 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.10 65.70 71.36 Avg2. 46451 46451 44940 96.75 96.75 96.75 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12112 11588 95.67 94.83 95.25 pp 4633 4677 4568 97.67 98.60 98.13 vp 4768 4733 4517 95.44 94.74 95.08 sbar 503 478 446 93.31 88.67 90.93 adjp 384 368 310 84.24 80.73 82.45 advp 822 832 695 83.53 84.55 84.04 prt 126 133 123 92.48 97.62 94.98 lst 10 3 3 100.00 30.00 46.15 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 6 85.71 37.50 52.17 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.81 70.72 76.29 Avg2. 23486 23343 22256 95.34 94.76 95.05 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 65 Log-likelihood = -156606.516401 Norm (log-likelihood gradient vector) = 17382.694422 Norm (lambda vector) = 129.474264 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4569 97.82 98.62 98.22 b-np 12220 12180 11929 97.94 97.62 97.78 i-np 13660 13854 13445 97.05 98.43 97.73 o 6349 6290 6144 97.68 96.77 97.22 b-vp 4768 4736 4623 97.61 96.96 97.29 i-vp 2602 2648 2514 94.94 96.62 95.77 b-adjp 384 368 317 86.14 82.55 84.31 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 829 707 85.28 86.01 85.65 i-advp 100 67 59 88.06 59.00 70.66 b-sbar 503 484 449 92.77 89.26 90.98 i-adjp 152 115 105 91.30 69.08 78.65 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 1 1 100.00 10.00 18.18 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.68 67.09 72.43 Avg2. 46451 46451 45054 96.99 96.99 96.99 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12180 11684 95.93 95.61 95.77 pp 4633 4671 4569 97.82 98.62 98.22 vp 4768 4736 4529 95.63 94.99 95.31 sbar 503 484 449 92.77 89.26 90.98 adjp 384 368 311 84.51 80.99 82.71 advp 822 829 695 83.84 84.55 84.19 prt 126 133 123 92.48 97.62 94.98 lst 10 1 1 100.00 10.00 18.18 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.30 70.79 76.54 Avg2. 23486 23412 22370 95.55 95.25 95.40 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 66 Log-likelihood = -154073.971818 Norm (log-likelihood gradient vector) = 8370.366034 Norm (lambda vector) = 128.237813 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4674 4567 97.71 98.58 98.14 b-np 12220 12197 11949 97.97 97.78 97.87 i-np 13660 13805 13436 97.33 98.36 97.84 o 6349 6315 6162 97.58 97.05 97.32 b-vp 4768 4741 4622 97.49 96.94 97.21 i-vp 2602 2653 2515 94.80 96.66 95.72 b-adjp 384 369 317 85.91 82.55 84.20 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 831 707 85.08 86.01 85.54 i-advp 100 69 59 85.51 59.00 69.82 b-sbar 503 480 448 93.33 89.07 91.15 i-adjp 152 113 103 91.15 67.76 77.74 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 8 7 87.50 43.75 58.33 i-conjp 24 15 14 93.33 58.33 71.79 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 1 1 100.00 10.00 18.18 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.35 65.62 71.42 Avg2. 46451 46451 45073 97.03 97.03 97.03 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12197 11717 96.06 95.88 95.97 pp 4633 4674 4567 97.71 98.58 98.14 vp 4768 4741 4527 95.49 94.95 95.22 sbar 503 480 448 93.33 89.07 91.15 adjp 384 369 310 84.01 80.73 82.34 advp 822 831 695 83.63 84.55 84.09 prt 126 134 124 92.54 98.41 95.38 lst 10 1 1 100.00 10.00 18.18 intj 4 0 0 0.00 0.00 0.00 conjp 16 8 7 87.50 43.75 58.33 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.03 69.59 75.72 Avg2. 23486 23435 22396 95.57 95.36 95.46 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 67 Log-likelihood = -152109.358137 Norm (log-likelihood gradient vector) = 4990.229831 Norm (lambda vector) = 128.568568 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4680 4568 97.61 98.60 98.10 b-np 12220 12166 11939 98.13 97.70 97.92 i-np 13660 13855 13460 97.15 98.54 97.84 o 6349 6291 6151 97.77 96.88 97.33 b-vp 4768 4740 4619 97.45 96.88 97.16 i-vp 2602 2664 2518 94.52 96.77 95.63 b-adjp 384 367 313 85.29 81.51 83.36 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 825 706 85.58 85.89 85.73 i-advp 100 67 58 86.57 58.00 69.46 b-sbar 503 478 446 93.31 88.67 90.93 i-adjp 152 115 102 88.70 67.11 76.40 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 7 6 85.71 37.50 52.17 i-conjp 24 13 12 92.31 50.00 64.86 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 3 3 100.00 30.00 46.15 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.12 65.72 71.39 Avg2. 46451 46451 45067 97.02 97.02 97.02 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12166 11707 96.23 95.80 96.01 pp 4633 4680 4568 97.61 98.60 98.10 vp 4768 4740 4522 95.40 94.84 95.12 sbar 503 478 446 93.31 88.67 90.93 adjp 384 367 306 83.38 79.69 81.49 advp 822 825 696 84.36 84.67 84.52 prt 126 134 124 92.54 98.41 95.38 lst 10 3 3 100.00 30.00 46.15 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 6 85.71 37.50 52.17 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.85 70.82 76.36 Avg2. 23486 23400 22378 95.63 95.28 95.46 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 68 Log-likelihood = -149916.244111 Norm (log-likelihood gradient vector) = 5305.847093 Norm (lambda vector) = 129.018425 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4567 97.61 98.58 98.09 b-np 12220 12215 11965 97.95 97.91 97.93 i-np 13660 13759 13423 97.56 98.27 97.91 o 6349 6329 6172 97.52 97.21 97.37 b-vp 4768 4744 4619 97.37 96.88 97.12 i-vp 2602 2667 2520 94.49 96.85 95.65 b-adjp 384 372 317 85.22 82.55 83.86 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 818 704 86.06 85.64 85.85 i-advp 100 66 58 87.88 58.00 69.88 b-sbar 503 479 448 93.53 89.07 91.24 i-adjp 152 120 106 88.33 69.74 77.94 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 7 6 85.71 37.50 52.17 i-conjp 24 13 12 92.31 50.00 64.86 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 3 3 100.00 30.00 46.15 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.20 65.93 71.54 Avg2. 46451 46451 45086 97.06 97.06 97.06 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12215 11745 96.15 96.11 96.13 pp 4633 4679 4567 97.61 98.58 98.09 vp 4768 4744 4522 95.32 94.84 95.08 sbar 503 479 448 93.53 89.07 91.24 adjp 384 372 310 83.33 80.73 82.01 advp 822 818 694 84.84 84.43 84.63 prt 126 134 124 92.54 98.41 95.38 lst 10 3 3 100.00 30.00 46.15 intj 4 0 0 0.00 0.00 0.00 conjp 16 7 6 85.71 37.50 52.17 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.90 70.97 76.47 Avg2. 23486 23451 22419 95.60 95.46 95.53 Current max chunk-based F1: 95.53 (iteration 55) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 69 Log-likelihood = -148867.214638 Norm (log-likelihood gradient vector) = 5301.284301 Norm (lambda vector) = 128.813344 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4568 97.63 98.60 98.11 b-np 12220 12174 11945 98.12 97.75 97.93 i-np 13660 13840 13457 97.23 98.51 97.87 o 6349 6290 6155 97.85 96.94 97.40 b-vp 4768 4747 4621 97.35 96.92 97.13 i-vp 2602 2661 2520 94.70 96.85 95.76 b-adjp 384 370 317 85.68 82.55 84.08 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 819 708 86.45 86.13 86.29 i-advp 100 65 58 89.23 58.00 70.30 b-sbar 503 477 446 93.50 88.67 91.02 i-adjp 152 123 110 89.43 72.37 80.00 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 8 7 87.50 43.75 58.33 i-conjp 24 15 14 93.33 58.33 71.79 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 3 3 100.00 30.00 46.15 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.52 66.79 72.18 Avg2. 46451 46451 45095 97.08 97.08 97.08 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12174 11719 96.26 95.90 96.08 pp 4633 4679 4568 97.63 98.60 98.11 vp 4768 4747 4527 95.37 94.95 95.16 sbar 503 477 446 93.50 88.67 91.02 adjp 384 370 311 84.05 80.99 82.49 advp 822 819 698 85.23 84.91 85.07 prt 126 134 124 92.54 98.41 95.38 lst 10 3 3 100.00 30.00 46.15 intj 4 0 0 0.00 0.00 0.00 conjp 16 8 7 87.50 43.75 58.33 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.21 71.62 76.98 Avg2. 23486 23411 22403 95.69 95.39 95.54 Current max chunk-based F1: 95.54 (iteration 69) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 70 Log-likelihood = -146970.025654 Norm (log-likelihood gradient vector) = 5963.316117 Norm (lambda vector) = 129.375804 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4676 4568 97.69 98.60 98.14 b-np 12220 12195 11957 98.05 97.85 97.95 i-np 13660 13796 13439 97.41 98.38 97.89 o 6349 6305 6161 97.72 97.04 97.38 b-vp 4768 4757 4625 97.23 97.00 97.11 i-vp 2602 2648 2514 94.94 96.62 95.77 b-adjp 384 370 318 85.95 82.81 84.35 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 823 711 86.39 86.50 86.44 i-advp 100 67 60 89.55 60.00 71.86 b-sbar 503 477 447 93.71 88.87 91.22 i-adjp 152 127 110 86.61 72.37 78.85 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 4 4 100.00 40.00 57.14 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.53 68.16 72.98 Avg2. 46451 46451 45104 97.10 97.10 97.10 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12195 11731 96.20 96.00 96.10 pp 4633 4676 4568 97.69 98.60 98.14 vp 4768 4757 4534 95.31 95.09 95.20 sbar 503 477 447 93.71 88.87 91.22 adjp 384 370 312 84.32 81.25 82.76 advp 822 823 702 85.30 85.40 85.35 prt 126 134 124 92.54 98.41 95.38 lst 10 4 4 100.00 40.00 57.14 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.40 73.36 78.06 Avg2. 23486 23445 22430 95.67 95.50 95.59 Current max chunk-based F1: 95.59 (iteration 70) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 71 Log-likelihood = -145136.364399 Norm (log-likelihood gradient vector) = 4149.874543 Norm (lambda vector) = 130.312755 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4565 97.90 98.53 98.21 b-np 12220 12199 11968 98.11 97.94 98.02 i-np 13660 13803 13455 97.48 98.50 97.99 o 6349 6307 6166 97.76 97.12 97.44 b-vp 4768 4753 4627 97.35 97.04 97.20 i-vp 2602 2646 2514 95.01 96.62 95.81 b-adjp 384 369 318 86.18 82.81 84.46 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 825 712 86.30 86.62 86.46 i-advp 100 67 60 89.55 60.00 71.86 b-sbar 503 483 451 93.37 89.66 91.48 i-adjp 152 127 110 86.61 72.37 78.85 b-prt 126 134 124 92.54 98.41 95.38 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 4 4 100.00 40.00 57.14 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.42 67.80 72.72 Avg2. 46451 46451 45139 97.18 97.18 97.18 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12199 11751 96.33 96.16 96.24 pp 4633 4663 4565 97.90 98.53 98.21 vp 4768 4753 4536 95.43 95.13 95.28 sbar 503 483 451 93.37 89.66 91.48 adjp 384 369 312 84.55 81.25 82.87 advp 822 825 703 85.21 85.52 85.37 prt 126 134 124 92.54 98.41 95.38 lst 10 4 4 100.00 40.00 57.14 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.42 73.47 78.13 Avg2. 23486 23439 22454 95.80 95.61 95.70 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 72 Log-likelihood = -143682.390484 Norm (log-likelihood gradient vector) = 4659.783963 Norm (lambda vector) = 131.190710 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4660 4565 97.96 98.53 98.25 b-np 12220 12241 11985 97.91 98.08 97.99 i-np 13660 13714 13407 97.76 98.15 97.95 o 6349 6350 6184 97.39 97.40 97.39 b-vp 4768 4742 4624 97.51 96.98 97.25 i-vp 2602 2646 2514 95.01 96.62 95.81 b-adjp 384 375 320 85.33 83.33 84.32 i-pp 52 35 34 97.14 65.38 78.16 b-advp 822 835 719 86.11 87.47 86.78 i-advp 100 68 61 89.71 61.00 72.62 b-sbar 503 484 451 93.18 89.66 91.39 i-adjp 152 128 110 85.94 72.37 78.57 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 4 4 100.00 40.00 57.14 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.32 67.88 72.73 Avg2. 46451 46451 45132 97.16 97.16 97.16 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12241 11765 96.11 96.28 96.19 pp 4633 4660 4565 97.96 98.53 98.25 vp 4768 4742 4532 95.57 95.05 95.31 sbar 503 484 451 93.18 89.66 91.39 adjp 384 375 314 83.73 81.77 82.74 advp 822 835 710 85.03 86.37 85.70 prt 126 133 123 92.48 97.62 94.98 lst 10 4 4 100.00 40.00 57.14 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.30 73.53 78.11 Avg2. 23486 23483 22472 95.69 95.68 95.69 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 73 Log-likelihood = -141381.484674 Norm (log-likelihood gradient vector) = 5696.865200 Norm (lambda vector) = 133.227751 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4662 4564 97.90 98.51 98.20 b-np 12220 12223 11977 97.99 98.01 98.00 i-np 13660 13743 13424 97.68 98.27 97.97 o 6349 6338 6179 97.49 97.32 97.41 b-vp 4768 4739 4622 97.53 96.94 97.23 i-vp 2602 2650 2514 94.87 96.62 95.73 b-adjp 384 374 320 85.56 83.33 84.43 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 834 718 86.09 87.35 86.71 i-advp 100 68 61 89.71 61.00 72.62 b-sbar 503 482 449 93.15 89.26 91.17 i-adjp 152 129 110 85.27 72.37 78.29 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 4 4 100.00 40.00 57.14 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.16 67.85 72.64 Avg2. 46451 46451 45130 97.16 97.16 97.16 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12223 11759 96.20 96.23 96.22 pp 4633 4662 4564 97.90 98.51 98.20 vp 4768 4739 4528 95.55 94.97 95.26 sbar 503 482 449 93.15 89.26 91.17 adjp 384 374 313 83.69 81.51 82.59 advp 822 834 709 85.01 86.25 85.63 prt 126 133 123 92.48 97.62 94.98 lst 10 4 4 100.00 40.00 57.14 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.29 73.44 78.05 Avg2. 23486 23460 22457 95.72 95.62 95.67 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 74 Log-likelihood = -139590.360696 Norm (log-likelihood gradient vector) = 4436.604558 Norm (lambda vector) = 134.348313 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4562 97.79 98.47 98.13 b-np 12220 12209 11963 97.99 97.90 97.94 i-np 13660 13760 13423 97.55 98.27 97.91 o 6349 6334 6173 97.46 97.23 97.34 b-vp 4768 4735 4620 97.57 96.90 97.23 i-vp 2602 2660 2519 94.70 96.81 95.74 b-adjp 384 370 318 85.95 82.81 84.35 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 826 715 86.56 86.98 86.77 i-advp 100 68 61 89.71 61.00 72.62 b-sbar 503 485 449 92.58 89.26 90.89 i-adjp 152 129 110 85.27 72.37 78.29 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.15 68.30 72.89 Avg2. 46451 46451 45106 97.10 97.10 97.10 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12209 11736 96.13 96.04 96.08 pp 4633 4665 4562 97.79 98.47 98.13 vp 4768 4735 4526 95.59 94.92 95.25 sbar 503 485 449 92.58 89.26 90.89 adjp 384 370 311 84.05 80.99 82.49 advp 822 826 706 85.47 85.89 85.68 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.30 74.32 78.55 Avg2. 23486 23437 22426 95.69 95.49 95.59 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 75 Log-likelihood = -137336.844826 Norm (log-likelihood gradient vector) = 4279.646395 Norm (lambda vector) = 136.014313 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4675 4568 97.71 98.60 98.15 b-np 12220 12218 11973 97.99 97.98 97.99 i-np 13660 13756 13423 97.58 98.27 97.92 o 6349 6343 6179 97.41 97.32 97.37 b-vp 4768 4734 4619 97.57 96.88 97.22 i-vp 2602 2661 2518 94.63 96.77 95.69 b-adjp 384 370 318 85.95 82.81 84.35 i-pp 52 37 34 91.89 65.38 76.40 b-advp 822 812 709 87.32 86.25 86.78 i-advp 100 68 61 89.71 61.00 72.62 b-sbar 503 482 450 93.36 89.46 91.37 i-adjp 152 122 107 87.70 70.39 78.10 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 9 7 77.78 58.33 66.67 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.60 68.19 73.03 Avg2. 46451 46451 45118 97.13 97.13 97.13 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12218 11744 96.12 96.10 96.11 pp 4633 4675 4568 97.71 98.60 98.15 vp 4768 4734 4525 95.59 94.90 95.24 sbar 503 482 450 93.36 89.46 91.37 adjp 384 370 311 84.05 80.99 82.49 advp 822 812 700 86.21 85.16 85.68 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.44 74.28 78.60 Avg2. 23486 23438 22434 95.72 95.52 95.62 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 76 Log-likelihood = -135855.943387 Norm (log-likelihood gradient vector) = 6645.624720 Norm (lambda vector) = 137.267575 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4565 97.69 98.53 98.11 b-np 12220 12194 11958 98.06 97.86 97.96 i-np 13660 13805 13443 97.38 98.41 97.89 o 6349 6318 6163 97.55 97.07 97.31 b-vp 4768 4737 4623 97.59 96.96 97.28 i-vp 2602 2659 2520 94.77 96.85 95.80 b-adjp 384 369 316 85.64 82.29 83.93 i-pp 52 37 34 91.89 65.38 76.40 b-advp 822 809 709 87.64 86.25 86.94 i-advp 100 68 61 89.71 61.00 72.62 b-sbar 503 484 449 92.77 89.26 90.98 i-adjp 152 126 107 84.92 70.39 76.98 b-prt 126 132 123 93.18 97.62 95.35 i-sbar 12 9 7 77.78 58.33 66.67 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.48 68.14 72.95 Avg2. 46451 46451 45107 97.11 97.11 97.11 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12194 11720 96.11 95.91 96.01 pp 4633 4673 4565 97.69 98.53 98.11 vp 4768 4737 4532 95.67 95.05 95.36 sbar 503 484 449 92.77 89.26 90.98 adjp 384 369 309 83.74 80.47 82.07 advp 822 809 700 86.53 85.16 85.84 prt 126 132 123 93.18 97.62 95.35 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.46 74.20 78.56 Avg2. 23486 23412 22411 95.72 95.42 95.57 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 77 Log-likelihood = -134123.892136 Norm (log-likelihood gradient vector) = 3454.596650 Norm (lambda vector) = 137.781184 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4677 4569 97.69 98.62 98.15 b-np 12220 12213 11971 98.02 97.96 97.99 i-np 13660 13766 13438 97.62 98.37 97.99 o 6349 6335 6178 97.52 97.31 97.41 b-vp 4768 4748 4628 97.47 97.06 97.27 i-vp 2602 2651 2516 94.91 96.69 95.79 b-adjp 384 370 316 85.41 82.29 83.82 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 806 706 87.59 85.89 86.73 i-advp 100 69 62 89.86 62.00 73.37 b-sbar 503 483 448 92.75 89.07 90.87 i-adjp 152 124 107 86.29 70.39 77.54 b-prt 126 132 123 93.18 97.62 95.35 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.29 68.18 72.89 Avg2. 46451 46451 45132 97.16 97.16 97.16 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12213 11750 96.21 96.15 96.18 pp 4633 4677 4569 97.69 98.62 98.15 vp 4768 4748 4538 95.58 95.18 95.38 sbar 503 483 448 92.75 89.07 90.87 adjp 384 370 309 83.51 80.47 81.96 advp 822 806 699 86.72 85.04 85.87 prt 126 132 123 93.18 97.62 95.35 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.45 74.21 78.56 Avg2. 23486 23443 22449 95.76 95.58 95.67 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 78 Log-likelihood = -132801.304877 Norm (log-likelihood gradient vector) = 3333.889728 Norm (lambda vector) = 138.174484 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4676 4572 97.78 98.68 98.23 b-np 12220 12174 11950 98.16 97.79 97.97 i-np 13660 13853 13469 97.23 98.60 97.91 o 6349 6289 6155 97.87 96.94 97.40 b-vp 4768 4752 4634 97.52 97.19 97.35 i-vp 2602 2643 2516 95.19 96.69 95.94 b-adjp 384 368 317 86.14 82.55 84.31 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 805 706 87.70 85.89 86.79 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 482 449 93.15 89.26 91.17 i-adjp 152 126 109 86.51 71.71 78.42 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.31 68.42 73.03 Avg2. 46451 46451 45135 97.17 97.17 97.17 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12174 11718 96.25 95.89 96.07 pp 4633 4676 4572 97.78 98.68 98.23 vp 4768 4752 4548 95.71 95.39 95.55 sbar 503 482 449 93.15 89.26 91.17 adjp 384 368 310 84.24 80.73 82.45 advp 822 805 701 87.08 85.28 86.17 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.56 74.29 78.65 Avg2. 23486 23404 22434 95.86 95.52 95.69 Current max chunk-based F1: 95.70 (iteration 71) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 79 Log-likelihood = -131137.712357 Norm (log-likelihood gradient vector) = 6154.151395 Norm (lambda vector) = 139.467993 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4677 4574 97.80 98.73 98.26 b-np 12220 12210 11973 98.06 97.98 98.02 i-np 13660 13775 13448 97.63 98.45 98.04 o 6349 6319 6174 97.71 97.24 97.47 b-vp 4768 4754 4634 97.48 97.19 97.33 i-vp 2602 2644 2515 95.12 96.66 95.88 b-adjp 384 370 318 85.95 82.81 84.35 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 806 707 87.72 86.01 86.86 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 484 450 92.98 89.46 91.19 i-adjp 152 128 109 85.16 71.71 77.86 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 9 8 88.89 50.00 64.00 i-conjp 24 17 16 94.12 66.67 78.05 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.24 68.51 73.05 Avg2. 46451 46451 45161 97.22 97.22 97.22 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12210 11749 96.22 96.15 96.19 pp 4633 4677 4574 97.80 98.73 98.26 vp 4768 4754 4547 95.65 95.36 95.51 sbar 503 484 450 92.98 89.46 91.19 adjp 384 370 311 84.05 80.99 82.49 advp 822 806 702 87.10 85.40 86.24 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 9 8 88.89 50.00 64.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.52 74.37 78.68 Avg2. 23486 23448 22469 95.82 95.67 95.75 Current max chunk-based F1: 95.75 (iteration 79) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 80 Log-likelihood = -129779.580992 Norm (log-likelihood gradient vector) = 3166.526057 Norm (lambda vector) = 140.300900 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4677 4573 97.78 98.70 98.24 b-np 12220 12228 11978 97.96 98.02 97.99 i-np 13660 13742 13424 97.69 98.27 97.98 o 6349 6328 6175 97.58 97.26 97.42 b-vp 4768 4744 4629 97.58 97.08 97.33 i-vp 2602 2656 2521 94.92 96.89 95.89 b-adjp 384 369 317 85.91 82.55 84.20 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 809 709 87.64 86.25 86.94 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 485 449 92.58 89.26 90.89 i-adjp 152 128 109 85.16 71.71 77.86 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.26 68.97 73.32 Avg2. 46451 46451 45144 97.19 97.19 97.19 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11752 96.11 96.17 96.14 pp 4633 4677 4573 97.78 98.70 98.24 vp 4768 4744 4542 95.74 95.26 95.50 sbar 503 485 449 92.58 89.26 90.89 adjp 384 369 310 84.01 80.73 82.34 advp 822 809 704 87.02 85.64 86.33 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.57 74.96 79.03 Avg2. 23486 23460 22467 95.77 95.66 95.71 Current max chunk-based F1: 95.75 (iteration 79) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 81 Log-likelihood = -127900.956395 Norm (log-likelihood gradient vector) = 3814.160233 Norm (lambda vector) = 141.877853 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4674 4571 97.80 98.66 98.23 b-np 12220 12230 11980 97.96 98.04 98.00 i-np 13660 13733 13418 97.71 98.23 97.97 o 6349 6334 6181 97.58 97.35 97.47 b-vp 4768 4738 4629 97.70 97.08 97.39 i-vp 2602 2665 2528 94.86 97.16 95.99 b-adjp 384 369 317 85.91 82.55 84.20 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 810 708 87.41 86.13 86.76 i-advp 100 74 65 87.84 65.00 74.71 b-sbar 503 484 450 92.98 89.46 91.19 i-adjp 152 126 109 86.51 71.71 78.42 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.34 69.51 73.66 Avg2. 46451 46451 45153 97.21 97.21 97.21 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12230 11753 96.10 96.18 96.14 pp 4633 4674 4571 97.80 98.66 98.23 vp 4768 4738 4542 95.86 95.26 95.56 sbar 503 484 450 92.98 89.46 91.19 adjp 384 369 310 84.01 80.73 82.34 advp 822 810 703 86.79 85.52 86.15 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.69 75.59 79.44 Avg2. 23486 23454 22467 95.79 95.66 95.73 Current max chunk-based F1: 95.75 (iteration 79) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 82 Log-likelihood = -126044.584179 Norm (log-likelihood gradient vector) = 4477.159750 Norm (lambda vector) = 143.171993 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4569 97.86 98.62 98.24 b-np 12220 12215 11976 98.04 98.00 98.02 i-np 13660 13774 13444 97.60 98.42 98.01 o 6349 6315 6173 97.75 97.23 97.49 b-vp 4768 4731 4626 97.78 97.02 97.40 i-vp 2602 2669 2529 94.75 97.19 95.96 b-adjp 384 367 316 86.10 82.29 84.15 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 812 712 87.68 86.62 87.15 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 486 450 92.59 89.46 91.00 i-adjp 152 123 107 86.99 70.39 77.82 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 10 7 70.00 58.33 63.64 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.50 70.07 74.05 Avg2. 46451 46451 45167 97.24 97.24 97.24 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12215 11759 96.27 96.23 96.25 pp 4633 4669 4569 97.86 98.62 98.24 vp 4768 4731 4536 95.88 95.13 95.50 sbar 503 486 450 92.59 89.46 91.00 adjp 384 367 309 84.20 80.47 82.29 advp 822 812 706 86.95 85.89 86.41 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.79 76.22 79.82 Avg2. 23486 23430 22468 95.89 95.67 95.78 Current max chunk-based F1: 95.78 (iteration 82) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 83 Log-likelihood = -122197.414334 Norm (log-likelihood gradient vector) = 5089.838351 Norm (lambda vector) = 145.567426 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4564 97.88 98.51 98.19 b-np 12220 12252 11971 97.71 97.96 97.83 i-np 13660 13683 13358 97.62 97.79 97.71 o 6349 6377 6184 96.97 97.40 97.19 b-vp 4768 4728 4623 97.78 96.96 97.37 i-vp 2602 2667 2527 94.75 97.12 95.92 b-adjp 384 366 320 87.43 83.33 85.33 i-pp 52 38 35 92.11 67.31 77.78 b-advp 822 812 713 87.81 86.74 87.27 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 490 453 92.45 90.06 91.24 i-adjp 152 119 107 89.92 70.39 78.97 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.52 70.96 74.55 Avg2. 46451 46451 45087 97.06 97.06 97.06 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12252 11724 95.69 95.94 95.82 pp 4633 4663 4564 97.88 98.51 98.19 vp 4768 4728 4532 95.85 95.05 95.45 sbar 503 490 453 92.45 90.06 91.24 adjp 384 366 313 85.52 81.51 83.47 advp 822 812 707 87.07 86.01 86.54 prt 126 134 123 91.79 97.62 94.62 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.79 76.35 79.90 Avg2. 23486 23462 22432 95.61 95.51 95.56 Current max chunk-based F1: 95.78 (iteration 82) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 84 Log-likelihood = -119907.293560 Norm (log-likelihood gradient vector) = 12356.418388 Norm (lambda vector) = 149.392871 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4566 97.88 98.55 98.21 b-np 12220 12243 11992 97.95 98.13 98.04 i-np 13660 13717 13412 97.78 98.18 97.98 o 6349 6352 6191 97.47 97.51 97.49 b-vp 4768 4729 4626 97.82 97.02 97.42 i-vp 2602 2666 2527 94.79 97.12 95.94 b-adjp 384 366 317 86.61 82.55 84.53 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 811 715 88.16 86.98 87.57 i-advp 100 73 66 90.41 66.00 76.30 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 122 107 87.70 70.39 78.10 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.91 71.02 74.75 Avg2. 46451 46451 45174 97.25 97.25 97.25 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12243 11773 96.16 96.34 96.25 pp 4633 4665 4566 97.88 98.55 98.21 vp 4768 4729 4535 95.90 95.11 95.50 sbar 503 488 453 92.83 90.06 91.42 adjp 384 366 310 84.70 80.73 82.67 advp 822 811 709 87.42 86.25 86.83 prt 126 133 123 92.48 97.62 94.98 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.90 76.34 79.94 Avg2. 23486 23452 22485 95.88 95.74 95.81 Current max chunk-based F1: 95.81 (iteration 84) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 85 Log-likelihood = -119841.781013 Norm (log-likelihood gradient vector) = 5519.063845 Norm (lambda vector) = 147.333466 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4565 97.90 98.53 98.21 b-np 12220 12214 11973 98.03 97.98 98.00 i-np 13660 13773 13429 97.50 98.31 97.90 o 6349 6333 6179 97.57 97.32 97.45 b-vp 4768 4731 4624 97.74 96.98 97.36 i-vp 2602 2662 2522 94.74 96.93 95.82 b-adjp 384 364 318 87.36 82.81 85.03 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 811 714 88.04 86.86 87.45 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 489 454 92.84 90.26 91.53 i-adjp 152 119 107 89.92 70.39 78.97 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 79.00 70.96 74.76 Avg2. 46451 46451 45152 97.20 97.20 97.20 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11748 96.18 96.14 96.16 pp 4633 4663 4565 97.90 98.53 98.21 vp 4768 4731 4532 95.79 95.05 95.42 sbar 503 489 454 92.84 90.26 91.53 adjp 384 364 311 85.44 80.99 83.16 advp 822 811 708 87.30 86.13 86.71 prt 126 134 123 91.79 97.62 94.62 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.89 76.35 79.94 Avg2. 23486 23423 22457 95.88 95.62 95.75 Current max chunk-based F1: 95.81 (iteration 84) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 86 Log-likelihood = -118243.752310 Norm (log-likelihood gradient vector) = 2900.696799 Norm (lambda vector) = 147.663590 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4565 97.84 98.53 98.18 b-np 12220 12190 11960 98.11 97.87 97.99 i-np 13660 13825 13454 97.32 98.49 97.90 o 6349 6314 6169 97.70 97.16 97.43 b-vp 4768 4733 4629 97.80 97.08 97.44 i-vp 2602 2657 2522 94.92 96.93 95.91 b-adjp 384 363 316 87.05 82.29 84.61 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 805 712 88.45 86.62 87.52 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 486 451 92.80 89.66 91.20 i-adjp 152 119 107 89.92 70.39 78.97 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.95 70.89 74.71 Avg2. 46451 46451 45152 97.20 97.20 97.20 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12190 11727 96.20 95.97 96.08 pp 4633 4666 4565 97.84 98.53 98.18 vp 4768 4733 4539 95.90 95.20 95.55 sbar 503 486 451 92.80 89.66 91.20 adjp 384 363 309 85.12 80.47 82.73 advp 822 805 705 87.58 85.77 86.66 prt 126 134 123 91.79 97.62 94.62 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.89 76.20 79.86 Avg2. 23486 23394 22435 95.90 95.52 95.71 Current max chunk-based F1: 95.81 (iteration 84) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 87 Log-likelihood = -117245.144355 Norm (log-likelihood gradient vector) = 3306.490815 Norm (lambda vector) = 147.569863 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4564 97.83 98.51 98.17 b-np 12220 12187 11955 98.10 97.83 97.96 i-np 13660 13838 13460 97.27 98.54 97.90 o 6349 6309 6168 97.77 97.15 97.46 b-vp 4768 4734 4631 97.82 97.13 97.47 i-vp 2602 2654 2523 95.06 96.96 96.00 b-adjp 384 362 315 87.02 82.03 84.45 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 803 710 88.42 86.37 87.38 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 485 450 92.78 89.46 91.09 i-adjp 152 121 107 88.43 70.39 78.39 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 5 5 100.00 50.00 66.67 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.65 70.86 74.55 Avg2. 46451 46451 45150 97.20 97.20 97.20 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12187 11727 96.23 95.97 96.10 pp 4633 4665 4564 97.83 98.51 98.17 vp 4768 4734 4541 95.92 95.24 95.58 sbar 503 485 450 92.78 89.46 91.09 adjp 384 362 308 85.08 80.21 82.57 advp 822 803 704 87.67 85.64 86.65 prt 126 134 123 91.79 97.62 94.62 lst 10 5 5 100.00 50.00 66.67 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.90 76.14 79.83 Avg2. 23486 23387 22433 95.92 95.52 95.72 Current max chunk-based F1: 95.81 (iteration 84) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 88 Log-likelihood = -116137.335247 Norm (log-likelihood gradient vector) = 3550.686751 Norm (lambda vector) = 147.953368 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4655 4560 97.96 98.42 98.19 b-np 12220 12082 11882 98.34 97.23 97.79 i-np 13660 14033 13511 96.28 98.91 97.58 o 6349 6211 6097 98.16 96.03 97.09 b-vp 4768 4733 4626 97.74 97.02 97.38 i-vp 2602 2659 2524 94.92 97.00 95.95 b-adjp 384 360 316 87.78 82.29 84.95 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 802 707 88.15 86.01 87.07 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 493 456 92.49 90.66 91.57 i-adjp 152 126 111 88.10 73.03 79.86 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 7 7 100.00 70.00 82.35 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.31 71.92 74.98 Avg2. 46451 46451 45058 97.00 97.00 97.00 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12082 11620 96.18 95.09 95.63 pp 4633 4655 4560 97.96 98.42 98.19 vp 4768 4733 4536 95.84 95.13 95.48 sbar 503 493 456 92.49 90.66 91.57 adjp 384 360 310 86.11 80.73 83.33 advp 822 802 701 87.41 85.28 86.33 prt 126 134 123 91.79 97.62 94.62 lst 10 7 7 100.00 70.00 82.35 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.94 78.17 80.95 Avg2. 23486 23278 22324 95.90 95.05 95.48 Current max chunk-based F1: 95.81 (iteration 84) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 89 Log-likelihood = -113705.751308 Norm (log-likelihood gradient vector) = 12988.602120 Norm (lambda vector) = 151.170608 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4657 4561 97.94 98.45 98.19 b-np 12220 12187 11965 98.18 97.91 98.05 i-np 13660 13809 13459 97.47 98.53 97.99 o 6349 6310 6173 97.83 97.23 97.53 b-vp 4768 4739 4632 97.74 97.15 97.44 i-vp 2602 2658 2525 95.00 97.04 96.01 b-adjp 384 362 316 87.29 82.29 84.72 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 809 712 88.01 86.62 87.31 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 494 456 92.31 90.66 91.47 i-adjp 152 128 111 86.72 73.03 79.29 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 7 7 100.00 70.00 82.35 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.88 71.41 74.50 Avg2. 46451 46451 45177 97.26 97.26 97.26 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12187 11749 96.41 96.15 96.28 pp 4633 4657 4561 97.94 98.45 98.19 vp 4768 4739 4543 95.86 95.28 95.57 sbar 503 494 455 92.11 90.46 91.27 adjp 384 362 310 85.64 80.73 83.11 advp 822 809 706 87.27 85.89 86.57 prt 126 134 123 91.79 97.62 94.62 lst 10 7 7 100.00 70.00 82.35 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.79 77.71 80.63 Avg2. 23486 23400 22464 96.00 95.65 95.82 Current max chunk-based F1: 95.82 (iteration 89) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 90 Log-likelihood = -111305.516777 Norm (log-likelihood gradient vector) = 3508.947711 Norm (lambda vector) = 153.007605 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4659 4563 97.94 98.49 98.21 b-np 12220 12213 11971 98.02 97.96 97.99 i-np 13660 13756 13429 97.62 98.31 97.96 o 6349 6330 6179 97.61 97.32 97.47 b-vp 4768 4738 4629 97.70 97.08 97.39 i-vp 2602 2662 2525 94.85 97.04 95.93 b-adjp 384 363 317 87.33 82.55 84.87 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 811 713 87.92 86.74 87.32 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 495 458 92.53 91.05 91.78 i-adjp 152 127 111 87.40 73.03 79.57 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 7 7 100.00 70.00 82.35 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.15 71.44 74.65 Avg2. 46451 46451 45162 97.23 97.23 97.23 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12213 11752 96.23 96.17 96.20 pp 4633 4659 4563 97.94 98.49 98.21 vp 4768 4738 4538 95.78 95.18 95.48 sbar 503 495 458 92.53 91.05 91.78 adjp 384 363 311 85.67 80.99 83.27 advp 822 811 707 87.18 86.01 86.59 prt 126 134 123 91.79 97.62 94.62 lst 10 7 7 100.00 70.00 82.35 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.80 77.80 80.69 Avg2. 23486 23431 22469 95.89 95.67 95.78 Current max chunk-based F1: 95.82 (iteration 89) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 91 Log-likelihood = -110862.956455 Norm (log-likelihood gradient vector) = 2665.912119 Norm (lambda vector) = 153.502232 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4565 97.90 98.53 98.21 b-np 12220 12231 11984 97.98 98.07 98.02 i-np 13660 13711 13407 97.78 98.15 97.96 o 6349 6341 6183 97.51 97.39 97.45 b-vp 4768 4733 4629 97.80 97.08 97.44 i-vp 2602 2666 2527 94.79 97.12 95.94 b-adjp 384 366 319 87.16 83.07 85.07 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 820 720 87.80 87.59 87.70 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 494 457 92.51 90.85 91.68 i-adjp 152 130 112 86.15 73.68 79.43 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 12 8 66.67 66.67 66.67 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 7 7 100.00 70.00 82.35 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.95 71.12 74.38 Avg2. 46451 46451 45169 97.24 97.24 97.24 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12231 11765 96.19 96.28 96.23 pp 4633 4663 4565 97.90 98.53 98.21 vp 4768 4733 4536 95.84 95.13 95.48 sbar 503 494 457 92.51 90.85 91.68 adjp 384 366 313 85.52 81.51 83.47 advp 822 820 714 87.07 86.86 86.97 prt 126 134 123 91.79 97.62 94.62 lst 10 7 7 100.00 70.00 82.35 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 83.77 77.93 80.75 Avg2. 23486 23459 22490 95.87 95.76 95.81 Current max chunk-based F1: 95.82 (iteration 89) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 92 Log-likelihood = -109997.740559 Norm (log-likelihood gradient vector) = 3395.798162 Norm (lambda vector) = 155.058511 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4568 97.79 98.60 98.19 b-np 12220 12232 11989 98.01 98.11 98.06 i-np 13660 13707 13411 97.84 98.18 98.01 o 6349 6345 6188 97.53 97.46 97.49 b-vp 4768 4728 4630 97.93 97.11 97.51 i-vp 2602 2665 2529 94.90 97.19 96.03 b-adjp 384 367 318 86.65 82.81 84.69 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 819 721 88.03 87.71 87.87 i-advp 100 76 67 88.16 67.00 76.14 b-sbar 503 489 454 92.84 90.26 91.53 i-adjp 152 128 110 85.94 72.37 78.57 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 8 61.54 66.67 64.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 9 9 100.00 90.00 94.74 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.70 72.09 74.79 Avg2. 46451 46451 45187 97.28 97.28 97.28 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11772 96.24 96.33 96.29 pp 4633 4671 4568 97.79 98.60 98.19 vp 4768 4728 4538 95.98 95.18 95.58 sbar 503 489 453 92.64 90.06 91.33 adjp 384 367 312 85.01 81.25 83.09 advp 822 819 715 87.30 86.98 87.14 prt 126 134 123 91.79 97.62 94.62 lst 10 9 8 88.89 80.00 84.21 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.66 78.85 80.71 Avg2. 23486 23460 22499 95.90 95.80 95.85 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 93 Log-likelihood = -108691.350313 Norm (log-likelihood gradient vector) = 3633.385672 Norm (lambda vector) = 157.129310 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4658 4558 97.85 98.38 98.12 b-np 12220 12228 11985 98.01 98.08 98.04 i-np 13660 13692 13403 97.89 98.12 98.00 o 6349 6350 6190 97.48 97.50 97.49 b-vp 4768 4733 4631 97.84 97.13 97.48 i-vp 2602 2666 2530 94.90 97.23 96.05 b-adjp 384 372 318 85.48 82.81 84.13 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 814 714 87.71 86.86 87.29 i-advp 100 79 66 83.54 66.00 73.74 b-sbar 503 504 458 90.87 91.05 90.96 i-adjp 152 135 112 82.96 73.68 78.05 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 8 61.54 66.67 64.00 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 16 15 93.75 62.50 75.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 9 9 100.00 90.00 94.74 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.06 71.26 74.05 Avg2. 46451 46451 45163 97.23 97.23 97.23 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11764 96.21 96.27 96.24 pp 4633 4658 4558 97.85 98.38 98.12 vp 4768 4733 4541 95.94 95.24 95.59 sbar 503 504 457 90.67 90.85 90.76 adjp 384 372 312 83.87 81.25 82.54 advp 822 814 706 86.73 85.89 86.31 prt 126 134 123 91.79 97.62 94.62 lst 10 9 8 88.89 80.00 84.21 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.20 78.18 80.14 Avg2. 23486 23462 22478 95.81 95.71 95.76 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 94 Log-likelihood = -108646.894253 Norm (log-likelihood gradient vector) = 12238.305112 Norm (lambda vector) = 162.985693 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4660 4562 97.90 98.47 98.18 b-np 12220 12234 11988 97.99 98.10 98.05 i-np 13660 13691 13400 97.87 98.10 97.99 o 6349 6353 6189 97.42 97.48 97.45 b-vp 4768 4729 4631 97.93 97.13 97.53 i-vp 2602 2666 2531 94.94 97.27 96.09 b-adjp 384 371 320 86.25 83.33 84.77 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 814 716 87.96 87.10 87.53 i-advp 100 77 66 85.71 66.00 74.58 b-sbar 503 497 457 91.95 90.85 91.40 i-adjp 152 135 112 82.96 73.68 78.05 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 8 61.54 66.67 64.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 9 9 100.00 90.00 94.74 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.37 72.12 74.65 Avg2. 46451 46451 45174 97.25 97.25 97.25 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12234 11768 96.19 96.30 96.25 pp 4633 4660 4562 97.90 98.47 98.18 vp 4768 4729 4542 96.05 95.26 95.65 sbar 503 497 456 91.75 90.66 91.20 adjp 384 371 314 84.64 81.77 83.18 advp 822 814 709 87.10 86.25 86.67 prt 126 134 123 91.79 97.62 94.62 lst 10 9 8 88.89 80.00 84.21 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.52 78.88 80.66 Avg2. 23486 23459 22492 95.88 95.77 95.82 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 95 Log-likelihood = -107671.458554 Norm (log-likelihood gradient vector) = 5290.946818 Norm (lambda vector) = 159.793819 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4568 97.75 98.60 98.17 b-np 12220 12219 11981 98.05 98.04 98.05 i-np 13660 13711 13414 97.83 98.20 98.02 o 6349 6349 6189 97.48 97.48 97.48 b-vp 4768 4730 4630 97.89 97.11 97.49 i-vp 2602 2661 2527 94.96 97.12 96.03 b-adjp 384 372 319 85.75 83.07 84.39 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 812 715 88.05 86.98 87.52 i-advp 100 79 66 83.54 66.00 73.74 b-sbar 503 490 455 92.86 90.46 91.64 i-adjp 152 134 112 83.58 73.68 78.32 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 10 9 90.00 56.25 69.23 i-conjp 24 16 15 93.75 62.50 75.00 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 9 9 100.00 90.00 94.74 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.35 71.67 74.41 Avg2. 46451 46451 45175 97.25 97.25 97.25 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12219 11761 96.25 96.24 96.25 pp 4633 4673 4568 97.75 98.60 98.17 vp 4768 4730 4540 95.98 95.22 95.60 sbar 503 490 454 92.65 90.26 91.44 adjp 384 372 313 84.14 81.51 82.80 advp 822 812 707 87.07 86.01 86.54 prt 126 134 123 91.79 97.62 94.62 lst 10 9 8 88.89 80.00 84.21 intj 4 0 0 0.00 0.00 0.00 conjp 16 10 9 90.00 56.25 69.23 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.45 78.17 80.25 Avg2. 23486 23449 22483 95.88 95.73 95.80 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 96 Log-likelihood = -105988.750406 Norm (log-likelihood gradient vector) = 2570.299168 Norm (lambda vector) = 162.773888 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4672 4568 97.77 98.60 98.18 b-np 12220 12208 11969 98.04 97.95 97.99 i-np 13660 13753 13433 97.67 98.34 98.00 o 6349 6336 6181 97.55 97.35 97.45 b-vp 4768 4730 4631 97.91 97.13 97.52 i-vp 2602 2655 2523 95.03 96.96 95.99 b-adjp 384 371 317 85.44 82.55 83.97 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 807 713 88.35 86.74 87.54 i-advp 100 77 66 85.71 66.00 74.58 b-sbar 503 486 452 93.00 89.86 91.41 i-adjp 152 132 110 83.33 72.37 77.46 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 9 9 100.00 90.00 94.74 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.54 72.26 74.80 Avg2. 46451 46451 45165 97.23 97.23 97.23 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12208 11751 96.26 96.16 96.21 pp 4633 4672 4568 97.77 98.60 98.18 vp 4768 4730 4538 95.94 95.18 95.56 sbar 503 486 451 92.80 89.66 91.20 adjp 384 371 311 83.83 80.99 82.38 advp 822 807 706 87.48 85.89 86.68 prt 126 134 123 91.79 97.62 94.62 lst 10 9 8 88.89 80.00 84.21 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.57 78.66 80.57 Avg2. 23486 23428 22466 95.89 95.66 95.78 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 97 Log-likelihood = -105025.196024 Norm (log-likelihood gradient vector) = 2564.206376 Norm (lambda vector) = 164.098711 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4675 4571 97.78 98.66 98.22 b-np 12220 12186 11958 98.13 97.86 97.99 i-np 13660 13821 13467 97.44 98.59 98.01 o 6349 6308 6169 97.80 97.16 97.48 b-vp 4768 4743 4637 97.77 97.25 97.51 i-vp 2602 2639 2516 95.34 96.69 96.01 b-adjp 384 369 314 85.09 81.77 83.40 i-pp 52 38 35 92.11 67.31 77.78 b-advp 822 795 708 89.06 86.13 87.57 i-advp 100 75 65 86.67 65.00 74.29 b-sbar 503 482 449 93.15 89.26 91.17 i-adjp 152 132 110 83.33 72.37 77.46 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 15 9 60.00 75.00 66.67 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.54 72.70 75.04 Avg2. 46451 46451 45168 97.24 97.24 97.24 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12186 11732 96.27 96.01 96.14 pp 4633 4675 4571 97.78 98.66 98.22 vp 4768 4743 4547 95.87 95.36 95.62 sbar 503 482 448 92.95 89.07 90.96 adjp 384 369 308 83.47 80.21 81.81 advp 822 795 701 88.18 85.28 86.70 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.72 78.47 80.06 Avg2. 23486 23405 22448 95.91 95.58 95.75 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 98 Log-likelihood = -103436.245794 Norm (log-likelihood gradient vector) = 7571.995856 Norm (lambda vector) = 167.897907 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4571 97.82 98.66 98.24 b-np 12220 12213 11973 98.03 97.98 98.01 i-np 13660 13767 13442 97.64 98.40 98.02 o 6349 6331 6184 97.68 97.40 97.54 b-vp 4768 4744 4636 97.72 97.23 97.48 i-vp 2602 2641 2517 95.30 96.73 96.01 b-adjp 384 369 316 85.64 82.29 83.93 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 795 706 88.81 85.89 87.32 i-advp 100 75 65 86.67 65.00 74.29 b-sbar 503 484 450 92.98 89.46 91.19 i-adjp 152 134 110 82.09 72.37 76.92 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 15 9 60.00 75.00 66.67 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.61 72.73 75.09 Avg2. 46451 46451 45174 97.25 97.25 97.25 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12213 11748 96.19 96.14 96.17 pp 4633 4673 4571 97.82 98.66 98.24 vp 4768 4744 4547 95.85 95.36 95.61 sbar 503 484 449 92.77 89.26 90.98 adjp 384 369 309 83.74 80.47 82.07 advp 822 795 699 87.92 85.04 86.46 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.70 78.51 80.07 Avg2. 23486 23433 22464 95.86 95.65 95.76 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 99 Log-likelihood = -102129.489438 Norm (log-likelihood gradient vector) = 3072.383593 Norm (lambda vector) = 168.662945 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4672 4571 97.84 98.66 98.25 b-np 12220 12220 11979 98.03 98.03 98.03 i-np 13660 13750 13431 97.68 98.32 98.00 o 6349 6335 6185 97.63 97.42 97.52 b-vp 4768 4746 4637 97.70 97.25 97.48 i-vp 2602 2642 2517 95.27 96.73 96.00 b-adjp 384 367 316 86.10 82.29 84.15 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 799 710 88.86 86.37 87.60 i-advp 100 75 65 86.67 65.00 74.29 b-sbar 503 487 452 92.81 89.86 91.31 i-adjp 152 133 110 82.71 72.37 77.19 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 15 9 60.00 75.00 66.67 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.65 72.78 75.14 Avg2. 46451 46451 45177 97.26 97.26 97.26 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12220 11749 96.15 96.15 96.15 pp 4633 4672 4571 97.84 98.66 98.25 vp 4768 4746 4547 95.81 95.36 95.59 sbar 503 487 451 92.61 89.66 91.11 adjp 384 367 309 84.20 80.47 82.29 advp 822 799 703 87.98 85.52 86.74 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.73 78.59 80.13 Avg2. 23486 23446 22471 95.84 95.68 95.76 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 100 Log-likelihood = -101208.702123 Norm (log-likelihood gradient vector) = 2647.327897 Norm (lambda vector) = 168.858168 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4572 97.90 98.68 98.29 b-np 12220 12268 12008 97.88 98.27 98.07 i-np 13660 13659 13388 98.02 98.01 98.01 o 6349 6373 6205 97.36 97.73 97.55 b-vp 4768 4739 4634 97.78 97.19 97.49 i-vp 2602 2653 2524 95.14 97.00 96.06 b-adjp 384 367 318 86.65 82.81 84.69 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 808 715 88.49 86.98 87.73 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 133 110 82.71 72.37 77.19 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 12 8 66.67 66.67 66.67 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.06 72.37 75.11 Avg2. 46451 46451 45195 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12268 11785 96.06 96.44 96.25 pp 4633 4670 4572 97.90 98.68 98.29 vp 4768 4739 4544 95.89 95.30 95.59 sbar 503 487 454 93.22 90.26 91.72 adjp 384 367 311 84.74 80.99 82.82 advp 822 808 708 87.62 86.13 86.87 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.81 78.79 80.27 Avg2. 23486 23494 22515 95.83 95.87 95.85 Current max chunk-based F1: 95.85 (iteration 92) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 101 Log-likelihood = -99516.004445 Norm (log-likelihood gradient vector) = 5255.808597 Norm (lambda vector) = 171.428395 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4569 97.96 98.62 98.29 b-np 12220 12244 12000 98.01 98.20 98.10 i-np 13660 13705 13413 97.87 98.19 98.03 o 6349 6353 6198 97.56 97.62 97.59 b-vp 4768 4736 4633 97.83 97.17 97.50 i-vp 2602 2661 2531 95.11 97.27 96.18 b-adjp 384 365 318 87.12 82.81 84.91 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 812 717 88.30 87.23 87.76 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 486 454 93.42 90.26 91.81 i-adjp 152 132 110 83.33 72.37 77.46 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.42 72.38 75.28 Avg2. 46451 46451 45209 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12244 11777 96.19 96.37 96.28 pp 4633 4664 4569 97.96 98.62 98.29 vp 4768 4736 4545 95.97 95.32 95.64 sbar 503 486 454 93.42 90.26 91.81 adjp 384 365 311 85.21 80.99 83.04 advp 822 812 710 87.44 86.37 86.90 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.89 78.81 80.32 Avg2. 23486 23462 22507 95.93 95.83 95.88 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 102 Log-likelihood = -98087.311668 Norm (log-likelihood gradient vector) = 2990.746284 Norm (lambda vector) = 172.491874 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4570 97.98 98.64 98.31 b-np 12220 12222 11987 98.08 98.09 98.09 i-np 13660 13754 13432 97.66 98.33 97.99 o 6349 6329 6187 97.76 97.45 97.60 b-vp 4768 4730 4627 97.82 97.04 97.43 i-vp 2602 2665 2532 95.01 97.31 96.15 b-adjp 384 366 318 86.89 82.81 84.80 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 813 716 88.07 87.10 87.58 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 482 450 93.36 89.46 91.37 i-adjp 152 133 110 82.71 72.37 77.19 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 11 8 72.73 66.67 69.57 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.36 72.33 75.22 Avg2. 46451 46451 45195 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12222 11761 96.23 96.24 96.24 pp 4633 4664 4570 97.98 98.64 98.31 vp 4768 4730 4538 95.94 95.18 95.56 sbar 503 482 450 93.36 89.46 91.37 adjp 384 366 311 84.97 80.99 82.93 advp 822 813 709 87.21 86.25 86.73 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.84 78.69 80.23 Avg2. 23486 23432 22480 95.94 95.72 95.83 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 103 Log-likelihood = -96500.634187 Norm (log-likelihood gradient vector) = 2715.769467 Norm (lambda vector) = 174.332887 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4574 98.05 98.73 98.39 b-np 12220 12211 11984 98.14 98.07 98.10 i-np 13660 13777 13444 97.58 98.42 98.00 o 6349 6320 6182 97.82 97.37 97.59 b-vp 4768 4729 4628 97.86 97.06 97.46 i-vp 2602 2664 2532 95.05 97.31 96.16 b-adjp 384 366 318 86.89 82.81 84.80 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 809 714 88.26 86.86 87.55 i-advp 100 73 63 86.30 63.00 72.83 b-sbar 503 481 451 93.76 89.66 91.67 i-adjp 152 133 110 82.71 72.37 77.19 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.17 72.75 75.36 Avg2. 46451 46451 45204 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12211 11758 96.29 96.22 96.25 pp 4633 4665 4574 98.05 98.73 98.39 vp 4768 4729 4540 96.00 95.22 95.61 sbar 503 481 451 93.76 89.66 91.67 adjp 384 366 311 84.97 80.99 82.93 advp 822 809 707 87.39 86.01 86.70 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.92 78.69 80.27 Avg2. 23486 23416 22482 96.01 95.73 95.87 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 104 Log-likelihood = -94448.554448 Norm (log-likelihood gradient vector) = 3220.561686 Norm (lambda vector) = 176.945300 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4674 4577 97.92 98.79 98.36 b-np 12220 12203 11982 98.19 98.05 98.12 i-np 13660 13771 13442 97.61 98.40 98.01 o 6349 6317 6184 97.89 97.40 97.65 b-vp 4768 4728 4630 97.93 97.11 97.51 i-vp 2602 2666 2533 95.01 97.35 96.17 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 810 714 88.15 86.86 87.50 i-advp 100 74 63 85.14 63.00 72.41 b-sbar 503 482 453 93.98 90.06 91.98 i-adjp 152 134 112 83.58 73.68 78.32 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.45 72.88 75.56 Avg2. 46451 46451 45215 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12203 11756 96.34 96.20 96.27 pp 4633 4674 4577 97.92 98.79 98.36 vp 4768 4728 4539 96.00 95.20 95.60 sbar 503 482 453 93.98 90.06 91.98 adjp 384 370 315 85.14 82.03 83.55 advp 822 810 706 87.16 85.89 86.52 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.92 78.83 80.35 Avg2. 23486 23422 22487 96.01 95.75 95.88 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 105 Log-likelihood = -91993.838349 Norm (log-likelihood gradient vector) = 3000.023490 Norm (lambda vector) = 179.522692 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4569 97.92 98.62 98.27 b-np 12220 12230 11990 98.04 98.12 98.08 i-np 13660 13728 13409 97.68 98.16 97.92 o 6349 6355 6197 97.51 97.61 97.56 b-vp 4768 4745 4646 97.91 97.44 97.68 i-vp 2602 2633 2525 95.90 97.04 96.47 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 802 708 88.28 86.13 87.19 i-advp 100 75 63 84.00 63.00 72.00 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 135 112 82.96 73.68 78.05 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.78 72.74 75.18 Avg2. 46451 46451 45196 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12230 11759 96.15 96.23 96.19 pp 4633 4666 4569 97.92 98.62 98.27 vp 4768 4745 4566 96.23 95.76 95.99 sbar 503 488 453 92.83 90.06 91.42 adjp 384 370 315 85.14 82.03 83.55 advp 822 802 700 87.28 85.16 86.21 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.82 78.80 80.28 Avg2. 23486 23456 22503 95.94 95.81 95.88 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 106 Log-likelihood = -88512.024888 Norm (log-likelihood gradient vector) = 7859.783891 Norm (lambda vector) = 186.332380 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4678 4574 97.78 98.73 98.25 b-np 12220 12220 11983 98.06 98.06 98.06 i-np 13660 13746 13422 97.64 98.26 97.95 o 6349 6328 6183 97.71 97.39 97.55 b-vp 4768 4743 4644 97.91 97.40 97.66 i-vp 2602 2636 2524 95.75 97.00 96.37 b-adjp 384 372 322 86.56 83.85 85.19 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 804 708 88.06 86.13 87.08 i-advp 100 74 62 83.78 62.00 71.26 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 138 114 82.61 75.00 78.62 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.73 72.76 75.16 Avg2. 46451 46451 45192 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12220 11750 96.15 96.15 96.15 pp 4633 4678 4574 97.78 98.73 98.25 vp 4768 4743 4561 96.16 95.66 95.91 sbar 503 488 453 92.83 90.06 91.42 adjp 384 372 317 85.22 82.55 83.86 advp 822 804 700 87.06 85.16 86.10 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.79 78.84 80.29 Avg2. 23486 23460 22496 95.89 95.78 95.84 Current max chunk-based F1: 95.88 (iteration 101) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 107 Log-likelihood = -86131.099731 Norm (log-likelihood gradient vector) = 2534.664181 Norm (lambda vector) = 187.731630 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4678 4575 97.80 98.75 98.27 b-np 12220 12219 11985 98.08 98.08 98.08 i-np 13660 13741 13425 97.70 98.28 97.99 o 6349 6330 6188 97.76 97.46 97.61 b-vp 4768 4746 4646 97.89 97.44 97.67 i-vp 2602 2634 2524 95.82 97.00 96.41 b-adjp 384 372 323 86.83 84.11 85.45 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 807 711 88.10 86.50 87.29 i-advp 100 75 63 84.00 63.00 72.00 b-sbar 503 488 454 93.03 90.26 91.62 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.80 72.86 75.25 Avg2. 46451 46451 45211 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12219 11758 96.23 96.22 96.22 pp 4633 4678 4575 97.80 98.75 98.27 vp 4768 4746 4564 96.17 95.72 95.94 sbar 503 488 454 93.03 90.26 91.62 adjp 384 372 318 85.48 82.81 84.13 advp 822 807 703 87.11 85.52 86.31 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.85 78.94 80.37 Avg2. 23486 23465 22513 95.94 95.86 95.90 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 108 Log-likelihood = -86177.268732 Norm (log-likelihood gradient vector) = 1919.039095 Norm (lambda vector) = 186.785730 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 319 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4678 4574 97.78 98.73 98.25 b-np 12220 12216 11982 98.08 98.05 98.07 i-np 13660 13751 13424 97.62 98.27 97.95 o 6349 6325 6182 97.74 97.37 97.55 b-vp 4768 4743 4644 97.91 97.40 97.66 i-vp 2602 2636 2524 95.75 97.00 96.37 b-adjp 384 373 322 86.33 83.85 85.07 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 805 709 88.07 86.25 87.15 i-advp 100 74 62 83.78 62.00 71.26 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 138 114 82.61 75.00 78.62 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.72 72.77 75.16 Avg2. 46451 46451 45193 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12216 11748 96.17 96.14 96.15 pp 4633 4678 4574 97.78 98.73 98.25 vp 4768 4743 4561 96.16 95.66 95.91 sbar 503 488 453 92.83 90.06 91.42 adjp 384 373 317 84.99 82.55 83.75 advp 822 805 701 87.08 85.28 86.17 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.77 78.85 80.29 Avg2. 23486 23458 22495 95.89 95.78 95.84 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 353 seconds Iteration: 109 Log-likelihood = -86121.637080 Norm (log-likelihood gradient vector) = 2351.750353 Norm (lambda vector) = 187.518540 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4679 4575 97.78 98.75 98.26 b-np 12220 12219 11985 98.08 98.08 98.08 i-np 13660 13741 13425 97.70 98.28 97.99 o 6349 6329 6187 97.76 97.45 97.60 b-vp 4768 4745 4645 97.89 97.42 97.66 i-vp 2602 2636 2524 95.75 97.00 96.37 b-adjp 384 372 323 86.83 84.11 85.45 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 807 711 88.10 86.50 87.29 i-advp 100 75 63 84.00 63.00 72.00 b-sbar 503 487 453 93.02 90.06 91.52 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.80 72.85 75.24 Avg2. 46451 46451 45208 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12219 11758 96.23 96.22 96.22 pp 4633 4679 4575 97.78 98.75 98.26 vp 4768 4745 4562 96.14 95.68 95.91 sbar 503 487 453 93.02 90.06 91.52 adjp 384 372 318 85.48 82.81 84.13 advp 822 807 703 87.11 85.52 86.31 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.85 78.92 80.35 Avg2. 23486 23464 22510 95.93 95.84 95.89 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 110 Log-likelihood = -86142.965631 Norm (log-likelihood gradient vector) = 2023.726993 Norm (lambda vector) = 187.034616 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4678 4574 97.78 98.73 98.25 b-np 12220 12216 11981 98.08 98.04 98.06 i-np 13660 13750 13423 97.62 98.27 97.94 o 6349 6325 6182 97.74 97.37 97.55 b-vp 4768 4744 4644 97.89 97.40 97.65 i-vp 2602 2636 2524 95.75 97.00 96.37 b-adjp 384 373 322 86.33 83.85 85.07 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 805 709 88.07 86.25 87.15 i-advp 100 74 62 83.78 62.00 71.26 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 138 114 82.61 75.00 78.62 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.72 72.77 75.16 Avg2. 46451 46451 45191 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12216 11747 96.16 96.13 96.15 pp 4633 4678 4574 97.78 98.73 98.25 vp 4768 4744 4561 96.14 95.66 95.90 sbar 503 488 453 92.83 90.06 91.42 adjp 384 373 317 84.99 82.55 83.75 advp 822 805 701 87.08 85.28 86.17 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.77 78.85 80.28 Avg2. 23486 23459 22494 95.89 95.78 95.83 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 111 Log-likelihood = -86121.168962 Norm (log-likelihood gradient vector) = 2271.986309 Norm (lambda vector) = 187.417294 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4672 4571 97.84 98.66 98.25 b-np 12220 12261 12006 97.92 98.25 98.08 i-np 13660 13663 13390 98.00 98.02 98.01 o 6349 6360 6201 97.50 97.67 97.58 b-vp 4768 4748 4642 97.77 97.36 97.56 i-vp 2602 2644 2526 95.54 97.08 96.30 b-adjp 384 368 321 87.23 83.59 85.37 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 814 720 88.45 87.59 88.02 i-advp 100 73 63 86.30 63.00 72.83 b-sbar 503 489 453 92.64 90.06 91.33 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 134 123 91.79 97.62 94.62 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 18 17 94.44 70.83 80.95 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.16 72.88 75.43 Avg2. 46451 46451 45210 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12261 11779 96.07 96.39 96.23 pp 4633 4672 4571 97.84 98.66 98.25 vp 4768 4748 4557 95.98 95.57 95.78 sbar 503 489 453 92.64 90.06 91.33 adjp 384 368 316 85.87 82.29 84.04 advp 822 814 712 87.47 86.62 87.04 prt 126 134 123 91.79 97.62 94.62 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.86 78.97 80.39 Avg2. 23486 23507 22529 95.84 95.93 95.88 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 112 Log-likelihood = -85767.216799 Norm (log-likelihood gradient vector) = 4498.998495 Norm (lambda vector) = 186.841668 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4569 97.94 98.62 98.28 b-np 12220 12188 11972 98.23 97.97 98.10 i-np 13660 13810 13458 97.45 98.52 97.98 o 6349 6292 6163 97.95 97.07 97.51 b-vp 4768 4748 4643 97.79 97.38 97.58 i-vp 2602 2648 2529 95.51 97.19 96.34 b-adjp 384 366 320 87.43 83.33 85.33 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 812 721 88.79 87.71 88.25 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 491 453 92.26 90.06 91.15 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.29 73.07 75.59 Avg2. 46451 46451 45209 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12188 11735 96.28 96.03 96.16 pp 4633 4665 4569 97.94 98.62 98.28 vp 4768 4748 4560 96.04 95.64 95.84 sbar 503 491 453 92.26 90.06 91.15 adjp 384 366 315 86.07 82.03 84.00 advp 822 812 713 87.81 86.74 87.27 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.98 78.92 80.42 Avg2. 23486 23424 22486 96.00 95.74 95.87 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 113 Log-likelihood = -85334.468517 Norm (log-likelihood gradient vector) = 4557.724770 Norm (lambda vector) = 187.248405 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4569 97.92 98.62 98.27 b-np 12220 12209 11982 98.14 98.05 98.10 i-np 13660 13762 13439 97.65 98.38 98.02 o 6349 6316 6178 97.82 97.31 97.56 b-vp 4768 4746 4641 97.79 97.34 97.56 i-vp 2602 2651 2529 95.40 97.19 96.29 b-adjp 384 367 321 87.47 83.59 85.49 i-pp 52 36 34 94.44 65.38 77.27 b-advp 822 812 720 88.67 87.59 88.13 i-advp 100 72 63 87.50 63.00 73.26 b-sbar 503 491 453 92.26 90.06 91.15 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.28 73.08 75.59 Avg2. 46451 46451 45213 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12209 11750 96.24 96.15 96.20 pp 4633 4666 4569 97.92 98.62 98.27 vp 4768 4746 4556 96.00 95.55 95.77 sbar 503 491 453 92.26 90.06 91.15 adjp 384 367 316 86.10 82.29 84.15 advp 822 812 712 87.68 86.62 87.15 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.96 78.94 80.42 Avg2. 23486 23445 22497 95.96 95.79 95.87 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 114 Log-likelihood = -84532.198243 Norm (log-likelihood gradient vector) = 2404.088137 Norm (lambda vector) = 188.306518 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4571 97.94 98.66 98.30 b-np 12220 12236 11996 98.04 98.17 98.10 i-np 13660 13700 13408 97.87 98.16 98.01 o 6349 6350 6196 97.57 97.59 97.58 b-vp 4768 4739 4638 97.87 97.27 97.57 i-vp 2602 2660 2530 95.11 97.23 96.16 b-adjp 384 364 320 87.91 83.33 85.56 i-pp 52 37 34 91.89 65.38 76.40 b-advp 822 814 719 88.33 87.47 87.90 i-advp 100 73 63 86.30 63.00 72.83 b-sbar 503 490 454 92.65 90.26 91.44 i-adjp 152 136 113 83.09 74.34 78.47 b-prt 126 132 122 92.42 96.83 94.57 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.09 73.01 75.47 Avg2. 46451 46451 45211 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12236 11767 96.17 96.29 96.23 pp 4633 4667 4570 97.92 98.64 98.28 vp 4768 4739 4549 95.99 95.41 95.70 sbar 503 490 454 92.65 90.26 91.44 adjp 384 364 315 86.54 82.03 84.22 advp 822 814 710 87.22 86.37 86.80 prt 126 132 122 92.42 96.83 94.57 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.98 78.83 80.38 Avg2. 23486 23463 22505 95.92 95.82 95.87 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 115 Log-likelihood = -83072.046439 Norm (log-likelihood gradient vector) = 2360.432335 Norm (lambda vector) = 190.833785 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4574 97.88 98.73 98.30 b-np 12220 12235 11996 98.05 98.17 98.11 i-np 13660 13699 13410 97.89 98.17 98.03 o 6349 6350 6197 97.59 97.61 97.60 b-vp 4768 4733 4632 97.87 97.15 97.51 i-vp 2602 2669 2530 94.79 97.23 96.00 b-adjp 384 364 320 87.91 83.33 85.56 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 813 720 88.56 87.59 88.07 i-advp 100 73 63 86.30 63.00 72.83 b-sbar 503 487 453 93.02 90.06 91.52 i-adjp 152 132 110 83.33 72.37 77.46 b-prt 126 132 122 92.42 96.83 94.57 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.00 72.91 75.37 Avg2. 46451 46451 45208 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12235 11769 96.19 96.31 96.25 pp 4633 4673 4573 97.86 98.70 98.28 vp 4768 4733 4539 95.90 95.20 95.55 sbar 503 487 453 93.02 90.06 91.52 adjp 384 364 314 86.26 81.77 83.96 advp 822 813 711 87.45 86.50 86.97 prt 126 132 122 92.42 96.83 94.57 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.00 78.79 80.36 Avg2. 23486 23458 22499 95.91 95.80 95.85 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 116 Log-likelihood = -81912.460834 Norm (log-likelihood gradient vector) = 2664.163454 Norm (lambda vector) = 192.813738 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4572 97.88 98.68 98.28 b-np 12220 12250 12001 97.97 98.21 98.09 i-np 13660 13665 13391 97.99 98.03 98.01 o 6349 6370 6203 97.38 97.70 97.54 b-vp 4768 4736 4636 97.89 97.23 97.56 i-vp 2602 2660 2531 95.15 97.27 96.20 b-adjp 384 375 325 86.67 84.64 85.64 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 801 713 89.01 86.74 87.86 i-advp 100 71 63 88.73 63.00 73.68 b-sbar 503 489 454 92.84 90.26 91.53 i-adjp 152 136 111 81.62 73.03 77.08 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.53 73.01 75.20 Avg2. 46451 46451 45204 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12250 11774 96.11 96.35 96.23 pp 4633 4671 4571 97.86 98.66 98.26 vp 4768 4736 4548 96.03 95.39 95.71 sbar 503 489 454 92.84 90.26 91.53 adjp 384 375 316 84.27 82.29 83.27 advp 822 801 705 88.01 85.77 86.88 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.85 78.88 80.34 Avg2. 23486 23476 22509 95.88 95.84 95.86 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 117 Log-likelihood = -80259.476432 Norm (log-likelihood gradient vector) = 3626.852920 Norm (lambda vector) = 194.976274 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4674 4573 97.84 98.70 98.27 b-np 12220 12220 11983 98.06 98.06 98.06 i-np 13660 13720 13416 97.78 98.21 98.00 o 6349 6333 6184 97.65 97.40 97.52 b-vp 4768 4732 4630 97.84 97.11 97.47 i-vp 2602 2674 2533 94.73 97.35 96.02 b-adjp 384 374 322 86.10 83.85 84.96 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 804 710 88.31 86.37 87.33 i-advp 100 73 64 87.67 64.00 73.99 b-sbar 503 484 452 93.39 89.86 91.59 i-adjp 152 135 110 81.48 72.37 76.66 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.57 73.46 75.46 Avg2. 46451 46451 45183 97.27 97.27 97.27 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12220 11758 96.22 96.22 96.22 pp 4633 4674 4572 97.82 98.68 98.25 vp 4768 4732 4538 95.90 95.18 95.54 sbar 503 484 452 93.39 89.86 91.59 adjp 384 374 313 83.69 81.51 82.59 advp 822 804 701 87.19 85.28 86.22 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.84 79.31 80.55 Avg2. 23486 23443 22476 95.88 95.70 95.79 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 118 Log-likelihood = -79700.669575 Norm (log-likelihood gradient vector) = 9024.115486 Norm (lambda vector) = 199.192306 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4570 97.86 98.64 98.25 b-np 12220 12237 11998 98.05 98.18 98.12 i-np 13660 13688 13407 97.95 98.15 98.05 o 6349 6355 6198 97.53 97.62 97.58 b-vp 4768 4737 4635 97.85 97.21 97.53 i-vp 2602 2665 2530 94.93 97.23 96.07 b-adjp 384 374 323 86.36 84.11 85.22 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 801 710 88.64 86.37 87.49 i-advp 100 71 63 88.73 63.00 73.68 b-sbar 503 489 454 92.84 90.26 91.53 i-adjp 152 137 110 80.29 72.37 76.12 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.42 72.93 75.11 Avg2. 46451 46451 45202 97.31 97.31 97.31 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12237 11777 96.24 96.37 96.31 pp 4633 4670 4569 97.84 98.62 98.23 vp 4768 4737 4545 95.95 95.32 95.63 sbar 503 489 454 92.84 90.26 91.53 adjp 384 374 314 83.96 81.77 82.85 advp 822 801 702 87.64 85.40 86.51 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.79 78.79 80.26 Avg2. 23486 23462 22502 95.91 95.81 95.86 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 119 Log-likelihood = -79204.577286 Norm (log-likelihood gradient vector) = 3738.930325 Norm (lambda vector) = 196.679687 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4572 97.88 98.68 98.28 b-np 12220 12233 11997 98.07 98.18 98.12 i-np 13660 13702 13416 97.91 98.21 98.06 o 6349 6343 6192 97.62 97.53 97.57 b-vp 4768 4735 4633 97.85 97.17 97.51 i-vp 2602 2663 2529 94.97 97.19 96.07 b-adjp 384 376 323 85.90 84.11 85.00 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 802 711 88.65 86.50 87.56 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 489 455 93.05 90.46 91.73 i-adjp 152 137 110 80.29 72.37 76.12 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 11 10 90.91 62.50 74.07 i-conjp 24 19 18 94.74 75.00 83.72 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.44 73.09 75.20 Avg2. 46451 46451 45207 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12233 11775 96.26 96.36 96.31 pp 4633 4671 4571 97.86 98.66 98.26 vp 4768 4735 4544 95.97 95.30 95.63 sbar 503 489 455 93.05 90.46 91.73 adjp 384 376 314 83.51 81.77 82.63 advp 822 802 703 87.66 85.52 86.58 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 11 10 90.91 62.50 74.07 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.77 78.82 80.27 Avg2. 23486 23460 22503 95.92 95.81 95.87 Current max chunk-based F1: 95.90 (iteration 107) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 120 Log-likelihood = -78315.861707 Norm (log-likelihood gradient vector) = 2300.284131 Norm (lambda vector) = 197.218365 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4573 97.99 98.70 98.34 b-np 12220 12213 11993 98.20 98.14 98.17 i-np 13660 13746 13439 97.77 98.38 98.07 o 6349 6316 6182 97.88 97.37 97.62 b-vp 4768 4733 4631 97.84 97.13 97.48 i-vp 2602 2665 2530 94.93 97.23 96.07 b-adjp 384 375 322 85.87 83.85 84.85 i-pp 52 39 36 92.31 69.23 79.12 b-advp 822 806 716 88.83 87.10 87.96 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 493 458 92.90 91.05 91.97 i-adjp 152 135 110 81.48 72.37 76.66 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.93 73.86 75.84 Avg2. 46451 46451 45228 97.37 97.37 97.37 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12213 11774 96.41 96.35 96.38 pp 4633 4667 4572 97.96 98.68 98.32 vp 4768 4733 4542 95.96 95.26 95.61 sbar 503 493 458 92.90 91.05 91.97 adjp 384 375 313 83.47 81.51 82.48 advp 822 806 708 87.84 86.13 86.98 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.87 79.54 80.69 Avg2. 23486 23442 22509 96.02 95.84 95.93 Current max chunk-based F1: 95.93 (iteration 120) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 121 Log-likelihood = -77246.693821 Norm (log-likelihood gradient vector) = 2171.737924 Norm (lambda vector) = 197.852397 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4575 97.97 98.75 98.36 b-np 12220 12213 11990 98.17 98.12 98.15 i-np 13660 13752 13441 97.74 98.40 98.07 o 6349 6307 6174 97.89 97.24 97.57 b-vp 4768 4736 4635 97.87 97.21 97.54 i-vp 2602 2663 2531 95.04 97.27 96.14 b-adjp 384 374 322 86.10 83.85 84.96 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 810 718 88.64 87.35 87.99 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 491 457 93.08 90.85 91.95 i-adjp 152 134 110 82.09 72.37 76.92 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.10 73.86 75.92 Avg2. 46451 46451 45227 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12213 11769 96.36 96.31 96.34 pp 4633 4670 4574 97.94 98.73 98.33 vp 4768 4736 4547 96.01 95.36 95.69 sbar 503 491 457 93.08 90.85 91.95 adjp 384 374 313 83.69 81.51 82.59 advp 822 810 710 87.65 86.37 87.01 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.89 79.55 80.70 Avg2. 23486 23449 22512 96.00 95.85 95.93 Current max chunk-based F1: 95.93 (iteration 120) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 122 Log-likelihood = -76381.001453 Norm (log-likelihood gradient vector) = 1965.589931 Norm (lambda vector) = 198.238641 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4574 97.99 98.73 98.36 b-np 12220 12172 11958 98.24 97.86 98.05 i-np 13660 13836 13463 97.30 98.56 97.93 o 6349 6265 6146 98.10 96.80 97.45 b-vp 4768 4737 4634 97.83 97.19 97.51 i-vp 2602 2669 2533 94.90 97.35 96.11 b-adjp 384 371 320 86.25 83.33 84.77 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 814 723 88.82 87.96 88.39 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 486 454 93.42 90.26 91.81 i-adjp 152 132 109 82.58 71.71 76.76 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.14 73.78 75.89 Avg2. 46451 46451 45188 97.28 97.28 97.28 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12172 11722 96.30 95.92 96.11 pp 4633 4668 4573 97.96 98.70 98.33 vp 4768 4737 4544 95.93 95.30 95.61 sbar 503 486 454 93.42 90.26 91.81 adjp 384 371 312 84.10 81.25 82.65 advp 822 814 715 87.84 86.98 87.41 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.97 79.48 80.71 Avg2. 23486 23403 22462 95.98 95.64 95.81 Current max chunk-based F1: 95.93 (iteration 120) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 123 Log-likelihood = -74407.921364 Norm (log-likelihood gradient vector) = 6148.687949 Norm (lambda vector) = 201.534786 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4576 97.99 98.77 98.38 b-np 12220 12238 12000 98.06 98.20 98.13 i-np 13660 13703 13415 97.90 98.21 98.05 o 6349 6328 6185 97.74 97.42 97.58 b-vp 4768 4738 4633 97.78 97.17 97.48 i-vp 2602 2668 2531 94.87 97.27 96.05 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 814 723 88.82 87.96 88.39 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 487 456 93.63 90.66 92.12 i-adjp 152 134 110 82.09 72.37 76.92 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.15 73.87 75.95 Avg2. 46451 46451 45224 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12238 11785 96.30 96.44 96.37 pp 4633 4670 4575 97.97 98.75 98.36 vp 4768 4738 4542 95.86 95.26 95.56 sbar 503 487 456 93.63 90.66 92.12 adjp 384 370 313 84.59 81.51 83.02 advp 822 814 715 87.84 86.98 87.41 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.03 79.60 80.80 Avg2. 23486 23472 22528 95.98 95.92 95.95 Current max chunk-based F1: 95.95 (iteration 123) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 124 Log-likelihood = -73326.436823 Norm (log-likelihood gradient vector) = 2065.910886 Norm (lambda vector) = 202.737574 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4576 97.92 98.77 98.35 b-np 12220 12253 12005 97.98 98.24 98.11 i-np 13660 13671 13401 98.03 98.10 98.06 o 6349 6347 6193 97.57 97.54 97.56 b-vp 4768 4738 4635 97.83 97.21 97.52 i-vp 2602 2666 2531 94.94 97.27 96.09 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 38 36 94.74 69.23 80.00 b-advp 822 813 723 88.93 87.96 88.44 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 484 453 93.60 90.06 91.79 i-adjp 152 135 111 82.22 73.03 77.35 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.16 73.88 75.96 Avg2. 46451 46451 45223 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12253 11790 96.22 96.48 96.35 pp 4633 4673 4575 97.90 98.75 98.32 vp 4768 4738 4543 95.88 95.28 95.58 sbar 503 484 453 93.60 90.06 91.79 adjp 384 370 313 84.59 81.51 83.02 advp 822 813 716 88.07 87.10 87.58 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.04 79.56 80.78 Avg2. 23486 23486 22532 95.94 95.94 95.94 Current max chunk-based F1: 95.95 (iteration 123) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 125 Log-likelihood = -72411.551527 Norm (log-likelihood gradient vector) = 2475.468386 Norm (lambda vector) = 204.433083 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4672 4574 97.90 98.73 98.31 b-np 12220 12250 11998 97.94 98.18 98.06 i-np 13660 13685 13402 97.93 98.11 98.02 o 6349 6349 6194 97.56 97.56 97.56 b-vp 4768 4738 4635 97.83 97.21 97.52 i-vp 2602 2659 2529 95.11 97.19 96.14 b-adjp 384 366 320 87.43 83.33 85.33 i-pp 52 37 35 94.59 67.31 78.65 b-advp 822 812 719 88.55 87.47 88.00 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 482 451 93.57 89.66 91.57 i-adjp 152 137 113 82.48 74.34 78.20 b-prt 126 133 123 92.48 97.62 94.98 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 20 19 95.00 79.17 86.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.12 73.78 75.89 Avg2. 46451 46451 45208 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12250 11778 96.15 96.38 96.26 pp 4633 4672 4573 97.88 98.70 98.29 vp 4768 4738 4545 95.93 95.32 95.62 sbar 503 482 451 93.57 89.66 91.57 adjp 384 366 313 85.52 81.51 83.47 advp 822 812 711 87.56 86.50 87.03 prt 126 133 123 92.48 97.62 94.98 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.08 79.44 80.74 Avg2. 23486 23475 22513 95.90 95.86 95.88 Current max chunk-based F1: 95.95 (iteration 123) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 126 Log-likelihood = -70623.333531 Norm (log-likelihood gradient vector) = 2630.417152 Norm (lambda vector) = 208.040624 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4675 4573 97.82 98.70 98.26 b-np 12220 12268 12005 97.86 98.24 98.05 i-np 13660 13657 13382 97.99 97.96 97.98 o 6349 6368 6204 97.42 97.72 97.57 b-vp 4768 4734 4637 97.95 97.25 97.60 i-vp 2602 2656 2530 95.26 97.23 96.23 b-adjp 384 365 320 87.67 83.33 85.45 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 809 719 88.88 87.47 88.17 i-advp 100 73 67 91.78 67.00 77.46 b-sbar 503 484 452 93.39 89.86 91.59 i-adjp 152 137 114 83.21 75.00 78.89 b-prt 126 130 122 93.85 96.83 95.31 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.19 73.95 76.01 Avg2. 46451 46451 45209 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12268 11783 96.05 96.42 96.23 pp 4633 4675 4572 97.80 98.68 98.24 vp 4768 4734 4549 96.09 95.41 95.75 sbar 503 484 452 93.39 89.86 91.59 adjp 384 365 313 85.75 81.51 83.58 advp 822 809 712 88.01 86.62 87.31 prt 126 130 122 93.85 96.83 95.31 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.26 79.41 80.81 Avg2. 23486 23487 22522 95.89 95.90 95.89 Current max chunk-based F1: 95.95 (iteration 123) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 127 Log-likelihood = -68464.639978 Norm (log-likelihood gradient vector) = 4059.726554 Norm (lambda vector) = 213.763948 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4676 4571 97.75 98.66 98.21 b-np 12220 12230 11991 98.05 98.13 98.09 i-np 13660 13727 13424 97.79 98.27 98.03 o 6349 6335 6188 97.68 97.46 97.57 b-vp 4768 4739 4639 97.89 97.29 97.59 i-vp 2602 2645 2527 95.54 97.12 96.32 b-adjp 384 369 320 86.72 83.33 84.99 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 804 719 89.43 87.47 88.44 i-advp 100 76 69 90.79 69.00 78.41 b-sbar 503 486 454 93.42 90.26 91.81 i-adjp 152 138 114 82.61 75.00 78.62 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.24 74.10 76.12 Avg2. 46451 46451 45223 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12230 11769 96.23 96.31 96.27 pp 4633 4676 4570 97.73 98.64 98.18 vp 4768 4739 4556 96.14 95.55 95.85 sbar 503 486 454 93.42 90.26 91.81 adjp 384 369 313 84.82 81.51 83.13 advp 822 804 713 88.68 86.74 87.70 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.26 79.54 80.88 Avg2. 23486 23457 22517 95.99 95.87 95.93 Current max chunk-based F1: 95.95 (iteration 123) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 128 Log-likelihood = -66615.440659 Norm (log-likelihood gradient vector) = 1856.085729 Norm (lambda vector) = 217.784017 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4674 4571 97.80 98.66 98.23 b-np 12220 12205 11980 98.16 98.04 98.10 i-np 13660 13763 13441 97.66 98.40 98.03 o 6349 6318 6179 97.80 97.32 97.56 b-vp 4768 4740 4638 97.85 97.27 97.56 i-vp 2602 2646 2527 95.50 97.12 96.30 b-adjp 384 371 321 86.52 83.59 85.03 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 803 718 89.41 87.35 88.37 i-advp 100 75 68 90.67 68.00 77.71 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 139 114 82.01 75.00 78.35 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.19 74.09 76.09 Avg2. 46451 46451 45222 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12205 11762 96.37 96.25 96.31 pp 4633 4674 4570 97.77 98.64 98.21 vp 4768 4740 4556 96.12 95.55 95.84 sbar 503 491 458 93.28 91.05 92.15 adjp 384 371 314 84.64 81.77 83.18 advp 822 803 712 88.67 86.62 87.63 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.24 79.63 80.91 Avg2. 23486 23437 22514 96.06 95.86 95.96 Current max chunk-based F1: 95.96 (iteration 128) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 129 Log-likelihood = -65991.982670 Norm (log-likelihood gradient vector) = 1774.481564 Norm (lambda vector) = 218.931764 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4570 97.80 98.64 98.22 b-np 12220 12202 11978 98.16 98.02 98.09 i-np 13660 13775 13442 97.58 98.40 97.99 o 6349 6305 6172 97.89 97.21 97.55 b-vp 4768 4740 4639 97.87 97.29 97.58 i-vp 2602 2648 2529 95.51 97.19 96.34 b-adjp 384 372 322 86.56 83.85 85.19 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 804 718 89.30 87.35 88.31 i-advp 100 75 68 90.67 68.00 77.71 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 139 114 82.01 75.00 78.35 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.32 74.24 76.23 Avg2. 46451 46451 45219 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12202 11759 96.37 96.23 96.30 pp 4633 4673 4569 97.77 98.62 98.19 vp 4768 4740 4558 96.16 95.60 95.88 sbar 503 491 458 93.28 91.05 92.15 adjp 384 372 315 84.68 82.03 83.33 advp 822 804 712 88.56 86.62 87.58 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.24 79.73 80.97 Avg2. 23486 23436 22514 96.07 95.86 95.96 Current max chunk-based F1: 95.96 (iteration 129) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 130 Log-likelihood = -65265.588349 Norm (log-likelihood gradient vector) = 2644.358980 Norm (lambda vector) = 220.520330 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4570 97.86 98.64 98.25 b-np 12220 12200 11978 98.18 98.02 98.10 i-np 13660 13769 13442 97.63 98.40 98.01 o 6349 6304 6172 97.91 97.21 97.56 b-vp 4768 4741 4641 97.89 97.34 97.61 i-vp 2602 2648 2529 95.51 97.19 96.34 b-adjp 384 376 324 86.17 84.38 85.26 i-pp 52 39 35 89.74 67.31 76.92 b-advp 822 807 720 89.22 87.59 88.40 i-advp 100 76 68 89.47 68.00 77.27 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 142 115 80.99 75.66 78.23 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 13 9 69.23 75.00 72.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.24 74.82 76.49 Avg2. 46451 46451 45226 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12200 11763 96.42 96.26 96.34 pp 4633 4670 4569 97.84 98.62 98.23 vp 4768 4741 4560 96.18 95.64 95.91 sbar 503 489 456 93.25 90.66 91.94 adjp 384 376 317 84.31 82.55 83.42 advp 822 807 714 88.48 86.86 87.66 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.27 80.40 81.33 Avg2. 23486 23438 22523 96.10 95.90 96.00 Current max chunk-based F1: 96.00 (iteration 130) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 131 Log-likelihood = -64380.966614 Norm (log-likelihood gradient vector) = 2731.111733 Norm (lambda vector) = 222.644667 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4652 4559 98.00 98.40 98.20 b-np 12220 12233 11988 98.00 98.10 98.05 i-np 13660 13716 13407 97.75 98.15 97.95 o 6349 6334 6179 97.55 97.32 97.44 b-vp 4768 4743 4641 97.85 97.34 97.59 i-vp 2602 2647 2528 95.50 97.16 96.32 b-adjp 384 378 324 85.71 84.38 85.04 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 808 721 89.23 87.71 88.47 i-advp 100 73 68 93.15 68.00 78.61 b-sbar 503 501 461 92.02 91.65 91.83 i-adjp 152 139 113 81.29 74.34 77.66 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 12 8 66.67 66.67 66.67 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.08 74.24 76.11 Avg2. 46451 46451 45197 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12233 11768 96.20 96.30 96.25 pp 4633 4652 4558 97.98 98.38 98.18 vp 4768 4743 4559 96.12 95.62 95.87 sbar 503 501 461 92.02 91.65 91.83 adjp 384 378 317 83.86 82.55 83.20 advp 822 808 716 88.61 87.10 87.85 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.10 80.42 81.25 Avg2. 23486 23469 22522 95.96 95.90 95.93 Current max chunk-based F1: 96.00 (iteration 130) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 132 Log-likelihood = -62363.382657 Norm (log-likelihood gradient vector) = 3196.002892 Norm (lambda vector) = 228.232692 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4659 4567 98.03 98.58 98.30 b-np 12220 12228 11991 98.06 98.13 98.09 i-np 13660 13740 13427 97.72 98.29 98.01 o 6349 6324 6181 97.74 97.35 97.55 b-vp 4768 4738 4638 97.89 97.27 97.58 i-vp 2602 2655 2531 95.33 97.27 96.29 b-adjp 384 371 320 86.25 83.33 84.77 i-pp 52 38 34 89.47 65.38 75.56 b-advp 822 811 719 88.66 87.47 88.06 i-advp 100 71 66 92.96 66.00 77.19 b-sbar 503 490 456 93.06 90.66 91.84 i-adjp 152 137 112 81.75 73.68 77.51 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 13 8 61.54 66.67 64.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.01 74.01 75.96 Avg2. 46451 46451 45216 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11775 96.30 96.36 96.33 pp 4633 4659 4566 98.00 98.55 98.28 vp 4768 4738 4554 96.12 95.51 95.81 sbar 503 490 455 92.86 90.46 91.64 adjp 384 371 313 84.37 81.51 82.91 advp 822 811 713 87.92 86.74 87.32 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.18 80.18 81.16 Avg2. 23486 23451 22519 96.03 95.88 95.95 Current max chunk-based F1: 96.00 (iteration 130) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 133 Log-likelihood = -61295.087442 Norm (log-likelihood gradient vector) = 2196.178552 Norm (lambda vector) = 229.685185 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4568 97.84 98.60 98.22 b-np 12220 12248 11997 97.95 98.18 98.06 i-np 13660 13696 13411 97.92 98.18 98.05 o 6349 6344 6187 97.53 97.45 97.49 b-vp 4768 4737 4637 97.89 97.25 97.57 i-vp 2602 2652 2528 95.32 97.16 96.23 b-adjp 384 373 322 86.33 83.85 85.07 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 812 720 88.67 87.59 88.13 i-advp 100 73 66 90.41 66.00 76.30 b-sbar 503 482 451 93.57 89.66 91.57 i-adjp 152 138 113 81.88 74.34 77.93 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.88 73.92 75.85 Avg2. 46451 46451 45207 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12248 11782 96.20 96.42 96.31 pp 4633 4669 4567 97.82 98.58 98.19 vp 4768 4737 4551 96.07 95.45 95.76 sbar 503 482 450 93.36 89.46 91.37 adjp 384 373 315 84.45 82.03 83.22 advp 822 812 713 87.81 86.74 87.27 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.13 79.50 80.79 Avg2. 23486 23474 22520 95.94 95.89 95.91 Current max chunk-based F1: 96.00 (iteration 130) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 134 Log-likelihood = -60677.620418 Norm (log-likelihood gradient vector) = 2496.487317 Norm (lambda vector) = 230.302579 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4570 97.88 98.64 98.26 b-np 12220 12245 12001 98.01 98.21 98.11 i-np 13660 13712 13420 97.87 98.24 98.06 o 6349 6340 6186 97.57 97.43 97.50 b-vp 4768 4737 4638 97.91 97.27 97.59 i-vp 2602 2651 2528 95.36 97.16 96.25 b-adjp 384 368 321 87.23 83.59 85.37 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 812 721 88.79 87.71 88.25 i-advp 100 74 67 90.54 67.00 77.01 b-sbar 503 482 451 93.57 89.66 91.57 i-adjp 152 134 112 83.58 73.68 78.32 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.03 73.93 75.93 Avg2. 46451 46451 45222 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12245 11785 96.24 96.44 96.34 pp 4633 4669 4569 97.86 98.62 98.24 vp 4768 4737 4553 96.12 95.49 95.80 sbar 503 482 450 93.36 89.46 91.37 adjp 384 368 314 85.33 81.77 83.51 advp 822 812 714 87.93 86.86 87.39 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.24 79.50 80.85 Avg2. 23486 23466 22527 96.00 95.92 95.96 Current max chunk-based F1: 96.00 (iteration 130) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 135 Log-likelihood = -60330.159281 Norm (log-likelihood gradient vector) = 1839.910976 Norm (lambda vector) = 229.820726 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4570 97.92 98.64 98.28 b-np 12220 12255 12009 97.99 98.27 98.13 i-np 13660 13687 13409 97.97 98.16 98.07 o 6349 6357 6198 97.50 97.62 97.56 b-vp 4768 4740 4641 97.91 97.34 97.62 i-vp 2602 2647 2527 95.47 97.12 96.29 b-adjp 384 370 326 88.11 84.90 86.47 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 805 715 88.82 86.98 87.89 i-advp 100 75 67 89.33 67.00 76.57 b-sbar 503 483 453 93.79 90.06 91.89 i-adjp 152 136 115 84.56 75.66 79.86 b-prt 126 131 123 93.89 97.62 95.72 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.99 74.19 76.04 Avg2. 46451 46451 45238 97.39 97.39 97.39 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12255 11795 96.25 96.52 96.38 pp 4633 4667 4569 97.90 98.62 98.26 vp 4768 4740 4557 96.14 95.57 95.86 sbar 503 483 452 93.58 89.86 91.68 adjp 384 370 320 86.49 83.33 84.88 advp 822 805 707 87.83 86.01 86.91 prt 126 131 123 93.89 97.62 95.72 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.37 79.63 80.98 Avg2. 23486 23473 22542 96.03 95.98 96.01 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 136 Log-likelihood = -59742.946970 Norm (log-likelihood gradient vector) = 2384.162187 Norm (lambda vector) = 229.513140 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4567 97.88 98.58 98.23 b-np 12220 12174 11966 98.29 97.92 98.11 i-np 13660 13833 13464 97.33 98.57 97.94 o 6349 6290 6158 97.90 96.99 97.44 b-vp 4768 4747 4642 97.79 97.36 97.57 i-vp 2602 2641 2524 95.57 97.00 96.28 b-adjp 384 367 324 88.28 84.38 86.28 i-pp 52 45 36 80.00 69.23 74.23 b-advp 822 801 714 89.14 86.86 87.99 i-advp 100 72 66 91.67 66.00 76.74 b-sbar 503 490 455 92.86 90.46 91.64 i-adjp 152 136 116 85.29 76.32 80.56 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.86 74.26 76.01 Avg2. 46451 46451 45206 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12174 11732 96.37 96.01 96.19 pp 4633 4666 4566 97.86 98.55 98.20 vp 4768 4747 4559 96.04 95.62 95.83 sbar 503 490 454 92.65 90.26 91.44 adjp 384 367 320 87.19 83.33 85.22 advp 822 801 708 88.39 86.13 87.25 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.41 79.71 81.04 Avg2. 23486 23399 22482 96.08 95.73 95.90 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 137 Log-likelihood = -58800.428971 Norm (log-likelihood gradient vector) = 6083.716823 Norm (lambda vector) = 232.763910 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4567 97.88 98.58 98.23 b-np 12220 12225 11994 98.11 98.15 98.13 i-np 13660 13730 13428 97.80 98.30 98.05 o 6349 6331 6183 97.66 97.39 97.52 b-vp 4768 4749 4642 97.75 97.36 97.55 i-vp 2602 2643 2524 95.50 97.00 96.24 b-adjp 384 368 325 88.32 84.64 86.44 i-pp 52 45 36 80.00 69.23 74.23 b-advp 822 805 715 88.82 86.98 87.89 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 490 455 92.86 90.46 91.64 i-adjp 152 136 116 85.29 76.32 80.56 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.72 74.29 75.97 Avg2. 46451 46451 45225 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12225 11773 96.30 96.34 96.32 pp 4633 4666 4566 97.86 98.55 98.20 vp 4768 4749 4558 95.98 95.60 95.79 sbar 503 490 454 92.65 90.26 91.44 adjp 384 368 321 87.23 83.59 85.37 advp 822 805 708 87.95 86.13 87.03 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.36 79.76 81.04 Avg2. 23486 23457 22523 96.02 95.90 95.96 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 138 Log-likelihood = -58272.051890 Norm (log-likelihood gradient vector) = 1915.593540 Norm (lambda vector) = 232.091574 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4568 97.84 98.60 98.22 b-np 12220 12242 11999 98.02 98.19 98.10 i-np 13660 13699 13414 97.92 98.20 98.06 o 6349 6340 6189 97.62 97.48 97.55 b-vp 4768 4749 4641 97.73 97.34 97.53 i-vp 2602 2644 2523 95.42 96.96 96.19 b-adjp 384 369 325 88.08 84.64 86.32 i-pp 52 45 36 80.00 69.23 74.23 b-advp 822 807 714 88.48 86.86 87.66 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 487 454 93.22 90.26 91.72 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.70 74.31 75.97 Avg2. 46451 46451 45220 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12242 11781 96.23 96.41 96.32 pp 4633 4669 4567 97.82 98.58 98.19 vp 4768 4749 4556 95.94 95.55 95.74 sbar 503 487 453 93.02 90.06 91.52 adjp 384 369 321 86.99 83.59 85.26 advp 822 807 707 87.61 86.01 86.80 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.32 79.74 81.01 Avg2. 23486 23477 22528 95.96 95.92 95.94 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 139 Log-likelihood = -57953.076402 Norm (log-likelihood gradient vector) = 1422.336175 Norm (lambda vector) = 232.613023 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4568 97.84 98.60 98.22 b-np 12220 12252 12005 97.98 98.24 98.11 i-np 13660 13674 13402 98.01 98.11 98.06 o 6349 6350 6192 97.51 97.53 97.52 b-vp 4768 4748 4640 97.73 97.32 97.52 i-vp 2602 2644 2523 95.42 96.96 96.19 b-adjp 384 371 326 87.87 84.90 86.36 i-pp 52 45 36 80.00 69.23 74.23 b-advp 822 808 716 88.61 87.10 87.85 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 488 455 93.24 90.46 91.83 i-adjp 152 138 117 84.78 76.97 80.69 b-prt 126 132 124 93.94 98.41 96.12 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.61 74.35 75.94 Avg2. 46451 46451 45220 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12252 11789 96.22 96.47 96.35 pp 4633 4669 4567 97.82 98.58 98.19 vp 4768 4748 4555 95.94 95.53 95.73 sbar 503 488 454 93.03 90.26 91.62 adjp 384 371 322 86.79 83.85 85.30 advp 822 808 708 87.62 86.13 86.87 prt 126 132 124 93.94 98.41 96.12 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.30 79.80 81.03 Avg2. 23486 23490 22538 95.95 95.96 95.96 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 140 Log-likelihood = -57246.813412 Norm (log-likelihood gradient vector) = 1848.473007 Norm (lambda vector) = 234.556148 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4568 97.86 98.60 98.23 b-np 12220 12242 12000 98.02 98.20 98.11 i-np 13660 13699 13413 97.91 98.19 98.05 o 6349 6334 6184 97.63 97.40 97.52 b-vp 4768 4744 4637 97.74 97.25 97.50 i-vp 2602 2651 2525 95.25 97.04 96.14 b-adjp 384 370 325 87.84 84.64 86.21 i-pp 52 42 35 83.33 67.31 74.47 b-advp 822 810 720 88.89 87.59 88.24 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 140 118 84.29 77.63 80.82 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.83 74.33 76.04 Avg2. 46451 46451 45221 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12242 11781 96.23 96.41 96.32 pp 4633 4668 4567 97.84 98.58 98.20 vp 4768 4744 4550 95.91 95.43 95.67 sbar 503 487 454 93.22 90.26 91.72 adjp 384 370 321 86.76 83.59 85.15 advp 822 810 713 88.02 86.74 87.38 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.36 79.90 81.11 Avg2. 23486 23476 22530 95.97 95.93 95.95 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 141 Log-likelihood = -56285.722273 Norm (log-likelihood gradient vector) = 2243.642980 Norm (lambda vector) = 236.963133 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4566 97.92 98.55 98.24 b-np 12220 12277 12015 97.87 98.32 98.09 i-np 13660 13625 13363 98.08 97.83 97.95 o 6349 6369 6197 97.30 97.61 97.45 b-vp 4768 4739 4634 97.78 97.19 97.49 i-vp 2602 2656 2527 95.14 97.12 96.12 b-adjp 384 373 326 87.40 84.90 86.13 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 811 721 88.90 87.71 88.30 i-advp 100 76 67 88.16 67.00 76.14 b-sbar 503 486 454 93.42 90.26 91.81 i-adjp 152 142 118 83.10 77.63 80.27 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 14 12 85.71 75.00 80.00 i-conjp 24 23 21 91.30 87.50 89.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.27 74.81 76.02 Avg2. 46451 46451 45199 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12277 11786 96.00 96.45 96.22 pp 4633 4663 4565 97.90 98.53 98.21 vp 4768 4739 4547 95.95 95.36 95.66 sbar 503 486 453 93.21 90.06 91.61 adjp 384 373 322 86.33 83.85 85.07 advp 822 811 715 88.16 86.98 87.57 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 14 12 85.71 75.00 80.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.72 80.54 81.13 Avg2. 23486 23506 22533 95.86 95.94 95.90 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 142 Log-likelihood = -55247.489939 Norm (log-likelihood gradient vector) = 4806.887019 Norm (lambda vector) = 241.599627 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4568 97.94 98.60 98.27 b-np 12220 12238 11999 98.05 98.19 98.12 i-np 13660 13711 13410 97.80 98.17 97.99 o 6349 6333 6182 97.62 97.37 97.49 b-vp 4768 4736 4635 97.87 97.21 97.54 i-vp 2602 2657 2529 95.18 97.19 96.18 b-adjp 384 371 325 87.60 84.64 86.09 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 807 720 89.22 87.59 88.40 i-advp 100 73 67 91.78 67.00 77.46 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 141 118 83.69 77.63 80.55 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.08 74.90 76.46 Avg2. 46451 46451 45220 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12238 11773 96.20 96.34 96.27 pp 4633 4664 4567 97.92 98.58 98.25 vp 4768 4736 4549 96.05 95.41 95.73 sbar 503 487 454 93.22 90.26 91.72 adjp 384 371 321 86.52 83.59 85.03 advp 822 807 714 88.48 86.86 87.66 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.47 80.52 81.48 Avg2. 23486 23459 22523 96.01 95.90 95.95 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 143 Log-likelihood = -54884.463572 Norm (log-likelihood gradient vector) = 1481.660152 Norm (lambda vector) = 240.638850 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4662 4567 97.96 98.58 98.27 b-np 12220 12230 12000 98.12 98.20 98.16 i-np 13660 13734 13429 97.78 98.31 98.04 o 6349 6319 6180 97.80 97.34 97.57 b-vp 4768 4735 4633 97.85 97.17 97.51 i-vp 2602 2656 2528 95.18 97.16 96.16 b-adjp 384 371 325 87.60 84.64 86.09 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 808 720 89.11 87.59 88.34 i-advp 100 73 67 91.78 67.00 77.46 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 141 118 83.69 77.63 80.55 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.08 74.91 76.46 Avg2. 46451 46451 45235 97.38 97.38 97.38 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12230 11782 96.34 96.42 96.38 pp 4633 4662 4566 97.94 98.55 98.25 vp 4768 4735 4547 96.03 95.36 95.70 sbar 503 489 455 93.05 90.46 91.73 adjp 384 371 321 86.52 83.59 85.03 advp 822 808 714 88.37 86.86 87.61 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.45 80.55 81.49 Avg2. 23486 23451 22530 96.07 95.93 96.00 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 144 Log-likelihood = -54806.629277 Norm (log-likelihood gradient vector) = 1386.246924 Norm (lambda vector) = 240.148208 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4659 4566 98.00 98.55 98.28 b-np 12220 12219 11992 98.14 98.13 98.14 i-np 13660 13761 13438 97.65 98.37 98.01 o 6349 6314 6175 97.80 97.26 97.53 b-vp 4768 4735 4633 97.85 97.17 97.51 i-vp 2602 2655 2526 95.14 97.08 96.10 b-adjp 384 370 327 88.38 85.16 86.74 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 808 720 89.11 87.59 88.34 i-advp 100 72 66 91.67 66.00 76.74 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.17 74.75 76.42 Avg2. 46451 46451 45227 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12219 11771 96.33 96.33 96.33 pp 4633 4659 4565 97.98 98.53 98.26 vp 4768 4735 4544 95.97 95.30 95.63 sbar 503 489 455 93.05 90.46 91.73 adjp 384 370 323 87.30 84.11 85.68 advp 822 808 714 88.37 86.86 87.61 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.53 80.58 81.54 Avg2. 23486 23436 22517 96.08 95.87 95.98 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 145 Log-likelihood = -54613.327830 Norm (log-likelihood gradient vector) = 1902.212145 Norm (lambda vector) = 240.047662 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4567 97.94 98.58 98.26 b-np 12220 12214 11992 98.18 98.13 98.16 i-np 13660 13769 13444 97.64 98.42 98.03 o 6349 6312 6174 97.81 97.24 97.53 b-vp 4768 4733 4633 97.89 97.17 97.53 i-vp 2602 2656 2526 95.11 97.08 96.08 b-adjp 384 372 326 87.63 84.90 86.24 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 807 719 89.10 87.47 88.28 i-advp 100 72 66 91.67 66.00 76.74 b-sbar 503 486 454 93.42 90.26 91.81 i-adjp 152 134 114 85.07 75.00 79.72 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.02 74.62 76.28 Avg2. 46451 46451 45226 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11773 96.39 96.34 96.37 pp 4633 4663 4566 97.92 98.55 98.24 vp 4768 4733 4543 95.99 95.28 95.63 sbar 503 486 453 93.21 90.06 91.61 adjp 384 372 321 86.29 83.59 84.92 advp 822 807 713 88.35 86.74 87.54 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.44 80.48 81.45 Avg2. 23486 23431 22514 96.09 95.86 95.97 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 146 Log-likelihood = -54264.670017 Norm (log-likelihood gradient vector) = 2293.672437 Norm (lambda vector) = 240.814029 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4566 97.84 98.55 98.19 b-np 12220 12246 12004 98.02 98.23 98.13 i-np 13660 13699 13417 97.94 98.22 98.08 o 6349 6348 6193 97.56 97.54 97.55 b-vp 4768 4735 4634 97.87 97.19 97.53 i-vp 2602 2648 2522 95.24 96.93 96.08 b-adjp 384 374 326 87.17 84.90 86.02 i-pp 52 43 34 79.07 65.38 71.58 b-advp 822 809 720 89.00 87.59 88.29 i-advp 100 74 67 90.54 67.00 77.01 b-sbar 503 485 454 93.61 90.26 91.90 i-adjp 152 131 112 85.50 73.68 79.15 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.77 74.61 76.16 Avg2. 46451 46451 45226 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12246 11789 96.27 96.47 96.37 pp 4633 4667 4565 97.81 98.53 98.17 vp 4768 4735 4544 95.97 95.30 95.63 sbar 503 485 453 93.40 90.06 91.70 adjp 384 374 320 85.56 83.33 84.43 advp 822 809 714 88.26 86.86 87.55 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.36 80.48 81.41 Avg2. 23486 23472 22530 95.99 95.93 95.96 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 147 Log-likelihood = -54153.393867 Norm (log-likelihood gradient vector) = 2352.212764 Norm (lambda vector) = 241.465610 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4567 97.86 98.58 98.22 b-np 12220 12229 11992 98.06 98.13 98.10 i-np 13660 13740 13425 97.71 98.28 97.99 o 6349 6322 6176 97.69 97.28 97.48 b-vp 4768 4740 4639 97.87 97.29 97.58 i-vp 2602 2645 2521 95.31 96.89 96.09 b-adjp 384 373 326 87.40 84.90 86.13 i-pp 52 44 34 77.27 65.38 70.83 b-advp 822 810 721 89.01 87.71 88.36 i-advp 100 72 67 93.06 67.00 77.91 b-sbar 503 488 457 93.65 90.85 92.23 i-adjp 152 129 112 86.82 73.68 79.72 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.89 74.63 76.23 Avg2. 46451 46451 45214 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12229 11770 96.25 96.32 96.28 pp 4633 4667 4566 97.84 98.55 98.19 vp 4768 4740 4549 95.97 95.41 95.69 sbar 503 488 456 93.44 90.66 92.03 adjp 384 373 320 85.79 83.33 84.54 advp 822 810 717 88.52 87.23 87.87 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.41 80.57 81.48 Avg2. 23486 23463 22523 95.99 95.90 95.95 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 148 Log-likelihood = -53463.954672 Norm (log-likelihood gradient vector) = 1962.531877 Norm (lambda vector) = 245.344176 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4569 97.86 98.62 98.24 b-np 12220 12236 11999 98.06 98.19 98.13 i-np 13660 13720 13425 97.85 98.28 98.06 o 6349 6331 6183 97.66 97.39 97.52 b-vp 4768 4743 4639 97.81 97.29 97.55 i-vp 2602 2647 2521 95.24 96.89 96.06 b-adjp 384 373 326 87.40 84.90 86.13 i-pp 52 42 34 80.95 65.38 72.34 b-advp 822 810 722 89.14 87.83 88.48 i-advp 100 73 67 91.78 67.00 77.46 b-sbar 503 486 455 93.62 90.46 92.01 i-adjp 152 129 112 86.82 73.68 79.72 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.01 74.63 76.28 Avg2. 46451 46451 45229 97.37 97.37 97.37 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12236 11784 96.31 96.43 96.37 pp 4633 4669 4568 97.84 98.60 98.22 vp 4768 4743 4548 95.89 95.39 95.64 sbar 503 486 454 93.42 90.26 91.81 adjp 384 373 320 85.79 83.33 84.54 advp 822 810 717 88.52 87.23 87.87 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.40 80.54 81.46 Avg2. 23486 23473 22536 96.01 95.96 95.98 Current max chunk-based F1: 96.01 (iteration 135) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 149 Log-likelihood = -53281.900548 Norm (log-likelihood gradient vector) = 1242.194927 Norm (lambda vector) = 244.647789 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4570 97.90 98.64 98.27 b-np 12220 12248 12008 98.04 98.27 98.15 i-np 13660 13690 13415 97.99 98.21 98.10 o 6349 6337 6189 97.66 97.48 97.57 b-vp 4768 4746 4640 97.77 97.32 97.54 i-vp 2602 2651 2526 95.28 97.08 96.17 b-adjp 384 374 328 87.70 85.42 86.54 i-pp 52 40 35 87.50 67.31 76.09 b-advp 822 814 725 89.07 88.20 88.63 i-advp 100 74 67 90.54 67.00 77.01 b-sbar 503 486 455 93.62 90.46 92.01 i-adjp 152 131 113 86.26 74.34 79.86 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.27 74.82 76.51 Avg2. 46451 46451 45248 97.41 97.41 97.41 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12248 11792 96.28 96.50 96.39 pp 4633 4668 4569 97.88 98.62 98.25 vp 4768 4746 4555 95.98 95.53 95.75 sbar 503 486 454 93.42 90.26 91.81 adjp 384 374 321 85.83 83.59 84.70 advp 822 814 719 88.33 87.47 87.90 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.40 80.62 81.50 Avg2. 23486 23492 22555 96.01 96.04 96.02 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 150 Log-likelihood = -53179.549904 Norm (log-likelihood gradient vector) = 1434.487762 Norm (lambda vector) = 243.977461 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4569 97.90 98.62 98.26 b-np 12220 12258 12010 97.98 98.28 98.13 i-np 13660 13669 13397 98.01 98.07 98.04 o 6349 6348 6191 97.53 97.51 97.52 b-vp 4768 4746 4641 97.79 97.34 97.56 i-vp 2602 2655 2530 95.29 97.23 96.25 b-adjp 384 373 326 87.40 84.90 86.13 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 812 722 88.92 87.83 88.37 i-advp 100 75 67 89.33 67.00 76.57 b-sbar 503 485 455 93.81 90.46 92.11 i-adjp 152 130 113 86.92 74.34 80.14 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.11 74.78 76.41 Avg2. 46451 46451 45233 97.38 97.38 97.38 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12258 11791 96.19 96.49 96.34 pp 4633 4667 4568 97.88 98.60 98.24 vp 4768 4746 4557 96.02 95.57 95.80 sbar 503 485 454 93.61 90.26 91.90 adjp 384 373 320 85.79 83.33 84.54 advp 822 812 716 88.18 87.10 87.64 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.40 80.56 81.47 Avg2. 23486 23497 22551 95.97 96.02 96.00 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 151 Log-likelihood = -53016.958900 Norm (log-likelihood gradient vector) = 2248.027752 Norm (lambda vector) = 244.402307 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4570 97.88 98.64 98.26 b-np 12220 12232 11996 98.07 98.17 98.12 i-np 13660 13711 13420 97.88 98.24 98.06 o 6349 6328 6184 97.72 97.40 97.56 b-vp 4768 4741 4636 97.79 97.23 97.51 i-vp 2602 2658 2530 95.18 97.23 96.20 b-adjp 384 372 325 87.37 84.64 85.98 i-pp 52 42 35 83.33 67.31 74.47 b-advp 822 815 721 88.47 87.71 88.09 i-advp 100 75 67 89.33 67.00 76.57 b-sbar 503 485 455 93.81 90.46 92.11 i-adjp 152 131 113 86.26 74.34 79.86 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.96 74.75 76.32 Avg2. 46451 46451 45229 97.37 97.37 97.37 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11779 96.30 96.39 96.34 pp 4633 4669 4569 97.86 98.62 98.24 vp 4768 4741 4550 95.97 95.43 95.70 sbar 503 485 454 93.61 90.26 91.90 adjp 384 372 320 86.02 83.33 84.66 advp 822 815 715 87.73 86.98 87.35 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.38 80.52 81.44 Avg2. 23486 23470 22532 96.00 95.94 95.97 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 152 Log-likelihood = -52648.719902 Norm (log-likelihood gradient vector) = 1606.201974 Norm (lambda vector) = 245.349202 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4567 97.92 98.58 98.25 b-np 12220 12260 12005 97.92 98.24 98.08 i-np 13660 13648 13375 98.00 97.91 97.96 o 6349 6362 6197 97.41 97.61 97.51 b-vp 4768 4742 4636 97.76 97.23 97.50 i-vp 2602 2657 2529 95.18 97.19 96.18 b-adjp 384 372 326 87.63 84.90 86.24 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 817 721 88.25 87.71 87.98 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 135 116 85.93 76.32 80.84 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.96 74.76 76.32 Avg2. 46451 46451 45204 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12260 11776 96.05 96.37 96.21 pp 4633 4664 4566 97.90 98.55 98.23 vp 4768 4742 4550 95.95 95.43 95.69 sbar 503 487 454 93.22 90.26 91.72 adjp 384 372 322 86.56 83.85 85.19 advp 822 817 716 87.64 87.10 87.37 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 80.58 81.43 Avg2. 23486 23499 22529 95.87 95.93 95.90 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 153 Log-likelihood = -52327.721927 Norm (log-likelihood gradient vector) = 3378.710306 Norm (lambda vector) = 247.141212 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4568 97.94 98.60 98.27 b-np 12220 12210 11982 98.13 98.05 98.09 i-np 13660 13751 13425 97.63 98.28 97.95 o 6349 6316 6176 97.78 97.28 97.53 b-vp 4768 4742 4637 97.79 97.25 97.52 i-vp 2602 2655 2527 95.18 97.12 96.14 b-adjp 384 370 324 87.57 84.38 85.94 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 815 720 88.34 87.59 87.97 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 487 456 93.63 90.66 92.12 i-adjp 152 134 115 85.82 75.66 80.42 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.98 74.69 76.30 Avg2. 46451 46451 45207 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12210 11754 96.27 96.19 96.23 pp 4633 4664 4567 97.92 98.58 98.25 vp 4768 4742 4550 95.95 95.43 95.69 sbar 503 487 455 93.43 90.46 91.92 adjp 384 370 320 86.49 83.33 84.88 advp 822 815 715 87.73 86.98 87.35 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.34 80.52 81.42 Avg2. 23486 23445 22506 95.99 95.83 95.91 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 154 Log-likelihood = -51874.677756 Norm (log-likelihood gradient vector) = 1809.444118 Norm (lambda vector) = 247.920888 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4570 97.90 98.64 98.27 b-np 12220 12214 11985 98.13 98.08 98.10 i-np 13660 13748 13425 97.65 98.28 97.96 o 6349 6315 6177 97.81 97.29 97.55 b-vp 4768 4740 4637 97.83 97.25 97.54 i-vp 2602 2655 2527 95.18 97.12 96.14 b-adjp 384 369 323 87.53 84.11 85.79 i-pp 52 42 35 83.33 67.31 74.47 b-advp 822 814 719 88.33 87.47 87.90 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 486 455 93.62 90.46 92.01 i-adjp 152 134 115 85.82 75.66 80.42 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.88 74.67 76.24 Avg2. 46451 46451 45210 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11761 96.29 96.24 96.27 pp 4633 4668 4569 97.88 98.62 98.25 vp 4768 4740 4550 95.99 95.43 95.71 sbar 503 486 454 93.42 90.26 91.81 adjp 384 369 319 86.45 83.07 84.73 advp 822 814 714 87.71 86.86 87.29 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.33 80.47 81.39 Avg2. 23486 23448 22512 96.01 95.85 95.93 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 155 Log-likelihood = -51655.064346 Norm (log-likelihood gradient vector) = 1490.705459 Norm (lambda vector) = 248.046739 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4568 97.90 98.60 98.25 b-np 12220 12211 11983 98.13 98.06 98.10 i-np 13660 13756 13431 97.64 98.32 97.98 o 6349 6307 6173 97.88 97.23 97.55 b-vp 4768 4743 4637 97.77 97.25 97.51 i-vp 2602 2655 2527 95.18 97.12 96.14 b-adjp 384 367 321 87.47 83.59 85.49 i-pp 52 43 35 81.40 67.31 73.68 b-advp 822 815 720 88.34 87.59 87.97 i-advp 100 73 65 89.04 65.00 75.14 b-sbar 503 488 455 93.24 90.46 91.83 i-adjp 152 134 115 85.82 75.66 80.42 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 9 64.29 75.00 69.23 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.76 74.64 76.17 Avg2. 46451 46451 45207 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12211 11759 96.30 96.23 96.26 pp 4633 4666 4567 97.88 98.58 98.23 vp 4768 4743 4551 95.95 95.45 95.70 sbar 503 488 454 93.03 90.26 91.62 adjp 384 367 317 86.38 82.55 84.42 advp 822 815 715 87.73 86.98 87.35 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 80.43 81.34 Avg2. 23486 23447 22508 96.00 95.84 95.92 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 156 Log-likelihood = -50922.707729 Norm (log-likelihood gradient vector) = 1600.651701 Norm (lambda vector) = 249.401659 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4565 97.79 98.53 98.16 b-np 12220 12259 12005 97.93 98.24 98.08 i-np 13660 13656 13382 97.99 97.96 97.98 o 6349 6355 6196 97.50 97.59 97.54 b-vp 4768 4746 4637 97.70 97.25 97.48 i-vp 2602 2646 2518 95.16 96.77 95.96 b-adjp 384 370 322 87.03 83.85 85.41 i-pp 52 45 35 77.78 67.31 72.16 b-advp 822 814 719 88.33 87.47 87.90 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 488 453 92.83 90.06 91.42 i-adjp 152 138 116 84.06 76.32 80.00 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.54 75.02 76.26 Avg2. 46451 46451 45190 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12259 11778 96.08 96.38 96.23 pp 4633 4668 4564 97.77 98.51 98.14 vp 4768 4746 4547 95.81 95.36 95.59 sbar 503 488 452 92.62 89.86 91.22 adjp 384 370 318 85.95 82.81 84.35 advp 822 814 714 87.71 86.86 87.29 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.15 80.40 81.27 Avg2. 23486 23502 22518 95.81 95.88 95.85 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 157 Log-likelihood = -50576.515924 Norm (log-likelihood gradient vector) = 3982.110467 Norm (lambda vector) = 252.061976 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4567 97.90 98.58 98.24 b-np 12220 12242 11994 97.97 98.15 98.06 i-np 13660 13696 13402 97.85 98.11 97.98 o 6349 6335 6187 97.66 97.45 97.56 b-vp 4768 4745 4638 97.74 97.27 97.51 i-vp 2602 2651 2524 95.21 97.00 96.10 b-adjp 384 370 323 87.30 84.11 85.68 i-pp 52 44 35 79.55 67.31 72.92 b-advp 822 812 719 88.55 87.47 88.00 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 136 116 85.29 76.32 80.56 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.74 75.07 76.39 Avg2. 46451 46451 45203 97.31 97.31 97.31 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12242 11767 96.12 96.29 96.21 pp 4633 4665 4566 97.88 98.55 98.21 vp 4768 4745 4550 95.89 95.43 95.66 sbar 503 489 455 93.05 90.46 91.73 adjp 384 370 319 86.22 83.07 84.62 advp 822 812 714 87.93 86.86 87.39 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.27 80.49 81.37 Avg2. 23486 23480 22516 95.89 95.87 95.88 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 158 Log-likelihood = -50509.552610 Norm (log-likelihood gradient vector) = 1690.209899 Norm (lambda vector) = 250.787679 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4566 97.88 98.55 98.21 b-np 12220 12237 11992 98.00 98.13 98.07 i-np 13660 13705 13404 97.80 98.13 97.96 o 6349 6330 6183 97.68 97.39 97.53 b-vp 4768 4745 4639 97.77 97.29 97.53 i-vp 2602 2649 2522 95.21 96.93 96.06 b-adjp 384 370 322 87.03 83.85 85.41 i-pp 52 44 35 79.55 67.31 72.92 b-advp 822 812 718 88.42 87.35 87.88 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 490 455 92.86 90.46 91.64 i-adjp 152 138 116 84.06 76.32 80.00 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.64 75.04 76.32 Avg2. 46451 46451 45194 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12237 11762 96.12 96.25 96.19 pp 4633 4665 4565 97.86 98.53 98.19 vp 4768 4745 4551 95.91 95.45 95.68 sbar 503 490 454 92.65 90.26 91.44 adjp 384 370 318 85.95 82.81 84.35 advp 822 812 713 87.81 86.74 87.27 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.19 80.43 81.30 Avg2. 23486 23476 22508 95.88 95.84 95.86 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 159 Log-likelihood = -50163.700631 Norm (log-likelihood gradient vector) = 1359.453858 Norm (lambda vector) = 251.597070 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4667 4567 97.86 98.58 98.22 b-np 12220 12235 11992 98.01 98.13 98.07 i-np 13660 13713 13409 97.78 98.16 97.97 o 6349 6323 6180 97.74 97.34 97.54 b-vp 4768 4748 4640 97.73 97.32 97.52 i-vp 2602 2651 2523 95.17 96.96 96.06 b-adjp 384 369 322 87.26 83.85 85.52 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 812 719 88.55 87.47 88.00 i-advp 100 72 64 88.89 64.00 74.42 b-sbar 503 487 453 93.02 90.06 91.52 i-adjp 152 137 116 84.67 76.32 80.28 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 14 12 85.71 75.00 80.00 i-conjp 24 23 21 91.30 87.50 89.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.33 74.93 76.11 Avg2. 46451 46451 45197 97.30 97.30 97.30 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12235 11761 96.13 96.24 96.18 pp 4633 4667 4566 97.84 98.55 98.19 vp 4768 4748 4552 95.87 95.47 95.67 sbar 503 487 452 92.81 89.86 91.31 adjp 384 369 318 86.18 82.81 84.46 advp 822 812 714 87.93 86.86 87.39 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 14 12 85.71 75.00 80.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.58 80.40 80.98 Avg2. 23486 23476 22508 95.88 95.84 95.86 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 160 Log-likelihood = -49491.293028 Norm (log-likelihood gradient vector) = 1125.656387 Norm (lambda vector) = 252.875661 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4660 4564 97.94 98.51 98.22 b-np 12220 12230 11995 98.08 98.16 98.12 i-np 13660 13706 13414 97.87 98.20 98.03 o 6349 6331 6186 97.71 97.43 97.57 b-vp 4768 4748 4643 97.79 97.38 97.58 i-vp 2602 2656 2528 95.18 97.16 96.16 b-adjp 384 371 323 87.06 84.11 85.56 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 816 722 88.48 87.83 88.16 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 491 455 92.67 90.46 91.55 i-adjp 152 135 114 84.44 75.00 79.44 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 14 12 85.71 75.00 80.00 i-conjp 24 23 21 91.30 87.50 89.36 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.47 74.93 76.18 Avg2. 46451 46451 45220 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12230 11775 96.28 96.36 96.32 pp 4633 4660 4563 97.92 98.49 98.20 vp 4768 4748 4557 95.98 95.57 95.78 sbar 503 491 454 92.46 90.26 91.35 adjp 384 371 318 85.71 82.81 84.24 advp 822 816 718 87.99 87.35 87.67 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 14 12 85.71 75.00 80.00 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.53 80.50 81.02 Avg2. 23486 23474 22530 95.98 95.93 95.95 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 161 Log-likelihood = -49034.316047 Norm (log-likelihood gradient vector) = 2663.668519 Norm (lambda vector) = 254.063319 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4660 4564 97.94 98.51 98.22 b-np 12220 12225 11992 98.09 98.13 98.11 i-np 13660 13717 13421 97.84 98.25 98.05 o 6349 6328 6185 97.74 97.42 97.58 b-vp 4768 4746 4641 97.79 97.34 97.56 i-vp 2602 2660 2531 95.15 97.27 96.20 b-adjp 384 370 324 87.57 84.38 85.94 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 813 719 88.44 87.47 87.95 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 492 456 92.68 90.66 91.66 i-adjp 152 134 115 85.82 75.66 80.42 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 13 11 84.62 68.75 75.86 i-conjp 24 22 20 90.91 83.33 86.96 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.38 74.45 75.89 Avg2. 46451 46451 45222 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12225 11771 96.29 96.33 96.31 pp 4633 4660 4563 97.92 98.49 98.20 vp 4768 4746 4556 96.00 95.55 95.77 sbar 503 492 455 92.48 90.46 91.46 adjp 384 370 319 86.22 83.07 84.62 advp 822 813 715 87.95 86.98 87.46 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 11 84.62 68.75 75.86 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.47 79.88 80.67 Avg2. 23486 23463 22523 95.99 95.90 95.95 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 162 Log-likelihood = -48588.926392 Norm (log-likelihood gradient vector) = 1188.876724 Norm (lambda vector) = 254.497938 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4566 97.88 98.55 98.21 b-np 12220 12225 11993 98.10 98.14 98.12 i-np 13660 13713 13422 97.88 98.26 98.07 o 6349 6334 6190 97.73 97.50 97.61 b-vp 4768 4744 4641 97.83 97.34 97.58 i-vp 2602 2660 2532 95.19 97.31 96.24 b-adjp 384 370 324 87.57 84.38 85.94 i-pp 52 40 33 82.50 63.46 71.74 b-advp 822 813 719 88.44 87.47 87.95 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 487 454 93.22 90.26 91.72 i-adjp 152 135 115 85.19 75.66 80.14 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 15 10 66.67 83.33 74.07 b-conjp 16 13 11 84.62 68.75 75.86 i-conjp 24 22 20 90.91 83.33 86.96 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.36 74.35 75.82 Avg2. 46451 46451 45229 97.37 97.37 97.37 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12225 11775 96.32 96.36 96.34 pp 4633 4665 4565 97.86 98.53 98.19 vp 4768 4744 4556 96.04 95.55 95.79 sbar 503 487 453 93.02 90.06 91.52 adjp 384 370 319 86.22 83.07 84.62 advp 822 813 715 87.95 86.98 87.46 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 11 84.62 68.75 75.86 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.53 79.85 80.68 Avg2. 23486 23461 22527 96.02 95.92 95.97 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 163 Log-likelihood = -48404.702741 Norm (log-likelihood gradient vector) = 966.019257 Norm (lambda vector) = 254.564931 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4567 97.84 98.58 98.20 b-np 12220 12214 11986 98.13 98.09 98.11 i-np 13660 13733 13427 97.77 98.29 98.03 o 6349 6328 6185 97.74 97.42 97.58 b-vp 4768 4739 4640 97.91 97.32 97.61 i-vp 2602 2656 2529 95.22 97.19 96.20 b-adjp 384 371 323 87.06 84.11 85.56 i-pp 52 42 34 80.95 65.38 72.34 b-advp 822 815 721 88.47 87.71 88.09 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 487 456 93.63 90.66 92.12 i-adjp 152 136 115 84.56 75.66 79.86 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.05 74.45 76.21 Avg2. 46451 46451 45223 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11767 96.34 96.29 96.32 pp 4633 4668 4566 97.81 98.55 98.18 vp 4768 4739 4554 96.10 95.51 95.80 sbar 503 487 455 93.43 90.46 91.92 adjp 384 371 318 85.71 82.81 84.24 advp 822 815 717 87.98 87.23 87.60 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.23 79.88 81.04 Avg2. 23486 23450 22521 96.04 95.89 95.96 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 164 Log-likelihood = -48234.500599 Norm (log-likelihood gradient vector) = 1698.609504 Norm (lambda vector) = 255.029763 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4662 4563 97.88 98.49 98.18 b-np 12220 12248 12002 97.99 98.22 98.10 i-np 13660 13658 13393 98.06 98.05 98.05 o 6349 6368 6202 97.39 97.68 97.54 b-vp 4768 4740 4644 97.97 97.40 97.69 i-vp 2602 2650 2528 95.40 97.16 96.27 b-adjp 384 376 326 86.70 84.90 85.79 i-pp 52 43 34 79.07 65.38 71.58 b-advp 822 817 724 88.62 88.08 88.35 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 135 113 83.70 74.34 78.75 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.90 74.52 76.17 Avg2. 46451 46451 45228 97.37 97.37 97.37 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12248 11785 96.22 96.44 96.33 pp 4633 4662 4562 97.85 98.47 98.16 vp 4768 4740 4562 96.24 95.68 95.96 sbar 503 491 457 93.08 90.85 91.95 adjp 384 376 320 85.11 83.33 84.21 advp 822 817 721 88.25 87.71 87.98 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.17 80.04 81.09 Avg2. 23486 23490 22551 96.00 96.02 96.01 Current max chunk-based F1: 96.02 (iteration 149) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 165 Log-likelihood = -48416.302071 Norm (log-likelihood gradient vector) = 3556.769936 Norm (lambda vector) = 255.774698 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4564 97.86 98.51 98.18 b-np 12220 12224 11993 98.11 98.14 98.13 i-np 13660 13713 13422 97.88 98.26 98.07 o 6349 6339 6192 97.68 97.53 97.60 b-vp 4768 4744 4646 97.93 97.44 97.69 i-vp 2602 2650 2529 95.43 97.19 96.31 b-adjp 384 372 324 87.10 84.38 85.71 i-pp 52 42 34 80.95 65.38 72.34 b-advp 822 816 723 88.60 87.96 88.28 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 136 115 84.56 75.66 79.86 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.06 74.49 76.23 Avg2. 46451 46451 45238 97.39 97.39 97.39 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12224 11778 96.35 96.38 96.37 pp 4633 4664 4563 97.83 98.49 98.16 vp 4768 4744 4565 96.23 95.74 95.98 sbar 503 489 455 93.05 90.46 91.73 adjp 384 372 319 85.75 83.07 84.39 advp 822 816 719 88.11 87.47 87.79 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.23 79.96 81.08 Avg2. 23486 23465 22543 96.07 95.98 96.03 Current max chunk-based F1: 96.03 (iteration 165) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 166 Log-likelihood = -48167.640805 Norm (log-likelihood gradient vector) = 1332.887387 Norm (lambda vector) = 255.250472 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4662 4563 97.88 98.49 98.18 b-np 12220 12227 11995 98.10 98.16 98.13 i-np 13660 13704 13418 97.91 98.23 98.07 o 6349 6344 6194 97.64 97.56 97.60 b-vp 4768 4741 4645 97.98 97.42 97.70 i-vp 2602 2648 2528 95.47 97.16 96.30 b-adjp 384 376 326 86.70 84.90 85.79 i-pp 52 43 34 79.07 65.38 71.58 b-advp 822 818 725 88.63 88.20 88.41 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 135 113 83.70 74.34 78.75 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.91 74.48 76.15 Avg2. 46451 46451 45239 97.39 97.39 97.39 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12227 11779 96.34 96.39 96.36 pp 4633 4662 4562 97.85 98.47 98.16 vp 4768 4741 4564 96.27 95.72 95.99 sbar 503 491 457 93.08 90.85 91.95 adjp 384 376 320 85.11 83.33 84.21 advp 822 818 721 88.14 87.71 87.93 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.17 80.04 81.09 Avg2. 23486 23471 22547 96.06 96.00 96.03 Current max chunk-based F1: 96.03 (iteration 166) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 167 Log-likelihood = -48049.039698 Norm (log-likelihood gradient vector) = 1174.993643 Norm (lambda vector) = 255.910392 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4566 97.92 98.55 98.24 b-np 12220 12256 12007 97.97 98.26 98.11 i-np 13660 13653 13387 98.05 98.00 98.03 o 6349 6364 6202 97.45 97.68 97.57 b-vp 4768 4744 4648 97.98 97.48 97.73 i-vp 2602 2647 2529 95.54 97.19 96.36 b-adjp 384 373 324 86.86 84.38 85.60 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 820 729 88.90 88.69 88.79 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 490 458 93.47 91.05 92.25 i-adjp 152 137 115 83.94 75.66 79.58 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.15 74.60 76.33 Avg2. 46451 46451 45240 97.39 97.39 97.39 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12256 11789 96.19 96.47 96.33 pp 4633 4663 4565 97.90 98.53 98.21 vp 4768 4744 4567 96.27 95.78 96.03 sbar 503 490 457 93.27 90.85 92.04 adjp 384 373 319 85.52 83.07 84.28 advp 822 820 726 88.54 88.32 88.43 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.26 80.10 81.17 Avg2. 23486 23502 22567 96.02 96.09 96.05 Current max chunk-based F1: 96.05 (iteration 167) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 168 Log-likelihood = -47865.523297 Norm (log-likelihood gradient vector) = 3245.663306 Norm (lambda vector) = 257.480128 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4567 97.90 98.58 98.24 b-np 12220 12233 12000 98.10 98.20 98.15 i-np 13660 13702 13416 97.91 98.21 98.06 o 6349 6335 6188 97.68 97.46 97.57 b-vp 4768 4744 4650 98.02 97.53 97.77 i-vp 2602 2646 2530 95.62 97.23 96.42 b-adjp 384 374 325 86.90 84.64 85.75 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 820 730 89.02 88.81 88.92 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 137 115 83.94 75.66 79.58 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.16 74.62 76.35 Avg2. 46451 46451 45254 97.42 97.42 97.42 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12233 11786 96.35 96.45 96.40 pp 4633 4665 4566 97.88 98.55 98.21 vp 4768 4744 4569 96.31 95.83 96.07 sbar 503 491 457 93.08 90.85 91.95 adjp 384 374 320 85.56 83.33 84.43 advp 822 820 727 88.66 88.44 88.55 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.28 80.14 81.20 Avg2. 23486 23483 22569 96.11 96.10 96.10 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 169 Log-likelihood = -47434.892190 Norm (log-likelihood gradient vector) = 1303.910402 Norm (lambda vector) = 258.299254 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4567 97.94 98.58 98.26 b-np 12220 12218 11995 98.17 98.16 98.17 i-np 13660 13730 13428 97.80 98.30 98.05 o 6349 6321 6181 97.79 97.35 97.57 b-vp 4768 4746 4648 97.94 97.48 97.71 i-vp 2602 2650 2530 95.47 97.23 96.34 b-adjp 384 372 323 86.83 84.11 85.45 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 819 728 88.89 88.56 88.73 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 491 458 93.28 91.05 92.15 i-adjp 152 137 115 83.94 75.66 79.58 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.14 74.58 76.32 Avg2. 46451 46451 45248 97.41 97.41 97.41 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12218 11783 96.44 96.42 96.43 pp 4633 4663 4566 97.92 98.55 98.24 vp 4768 4746 4567 96.23 95.78 96.01 sbar 503 491 457 93.08 90.85 91.95 adjp 384 372 318 85.48 82.81 84.13 advp 822 819 725 88.52 88.20 88.36 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.26 80.06 81.15 Avg2. 23486 23465 22560 96.14 96.06 96.10 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 170 Log-likelihood = -47075.716385 Norm (log-likelihood gradient vector) = 1071.205033 Norm (lambda vector) = 259.002436 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4568 97.96 98.60 98.28 b-np 12220 12214 11992 98.18 98.13 98.16 i-np 13660 13739 13433 97.77 98.34 98.05 o 6349 6321 6183 97.82 97.39 97.60 b-vp 4768 4744 4644 97.89 97.40 97.65 i-vp 2602 2651 2528 95.36 97.16 96.25 b-adjp 384 371 322 86.79 83.85 85.30 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 818 727 88.88 88.44 88.66 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 490 457 93.27 90.85 92.04 i-adjp 152 137 115 83.94 75.66 79.58 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.24 74.54 76.34 Avg2. 46451 46451 45244 97.40 97.40 97.40 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11781 96.45 96.41 96.43 pp 4633 4663 4567 97.94 98.58 98.26 vp 4768 4744 4558 96.08 95.60 95.84 sbar 503 490 456 93.06 90.66 91.84 adjp 384 371 317 85.44 82.55 83.97 advp 822 818 724 88.51 88.08 88.29 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.24 79.98 81.10 Avg2. 23486 23456 22547 96.12 96.00 96.06 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 171 Log-likelihood = -46652.304445 Norm (log-likelihood gradient vector) = 1194.781265 Norm (lambda vector) = 259.756369 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4569 97.92 98.62 98.27 b-np 12220 12216 11988 98.13 98.10 98.12 i-np 13660 13741 13433 97.76 98.34 98.05 o 6349 6322 6182 97.79 97.37 97.58 b-vp 4768 4740 4640 97.89 97.32 97.60 i-vp 2602 2651 2527 95.32 97.12 96.21 b-adjp 384 371 322 86.79 83.85 85.30 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 819 726 88.64 88.32 88.48 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 485 455 93.81 90.46 92.11 i-adjp 152 136 115 84.56 75.66 79.86 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 14 10 71.43 83.33 76.92 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.43 75.03 76.69 Avg2. 46451 46451 45234 97.38 97.38 97.38 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12216 11775 96.39 96.36 96.37 pp 4633 4666 4568 97.90 98.60 98.25 vp 4768 4740 4554 96.08 95.51 95.79 sbar 503 485 454 93.61 90.26 91.90 adjp 384 371 317 85.44 82.55 83.97 advp 822 819 723 88.28 87.96 88.12 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.33 80.54 81.43 Avg2. 23486 23454 22536 96.09 95.96 96.02 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 172 Log-likelihood = -45676.348104 Norm (log-likelihood gradient vector) = 1460.865787 Norm (lambda vector) = 261.882559 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4655 4563 98.02 98.49 98.26 b-np 12220 12222 11986 98.07 98.09 98.08 i-np 13660 13725 13419 97.77 98.24 98.00 o 6349 6337 6185 97.60 97.42 97.51 b-vp 4768 4734 4636 97.93 97.23 97.58 i-vp 2602 2654 2528 95.25 97.16 96.19 b-adjp 384 372 323 86.83 84.11 85.45 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 818 723 88.39 87.96 88.17 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 497 461 92.76 91.65 92.20 i-adjp 152 135 115 85.19 75.66 80.14 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.73 75.07 76.85 Avg2. 46451 46451 45216 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12222 11765 96.26 96.28 96.27 pp 4633 4655 4562 98.00 98.47 98.23 vp 4768 4734 4549 96.09 95.41 95.75 sbar 503 497 461 92.76 91.65 92.20 adjp 384 372 318 85.48 82.81 84.13 advp 822 818 720 88.02 87.59 87.80 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.22 80.64 81.42 Avg2. 23486 23455 22520 96.01 95.89 95.95 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 173 Log-likelihood = -44840.991464 Norm (log-likelihood gradient vector) = 2331.843343 Norm (lambda vector) = 264.081039 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4567 97.94 98.58 98.26 b-np 12220 12223 11989 98.09 98.11 98.10 i-np 13660 13725 13423 97.80 98.27 98.03 o 6349 6335 6186 97.65 97.43 97.54 b-vp 4768 4735 4636 97.91 97.23 97.57 i-vp 2602 2652 2527 95.29 97.12 96.19 b-adjp 384 371 322 86.79 83.85 85.30 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 819 724 88.40 88.08 88.24 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 492 460 93.50 91.45 92.46 i-adjp 152 135 115 85.19 75.66 80.14 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.61 74.54 76.52 Avg2. 46451 46451 45224 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12223 11771 96.30 96.33 96.31 pp 4633 4663 4566 97.92 98.55 98.24 vp 4768 4735 4549 96.07 95.41 95.74 sbar 503 492 460 93.50 91.45 92.46 adjp 384 371 317 85.44 82.55 83.97 advp 822 819 721 88.03 87.71 87.87 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.22 80.00 81.09 Avg2. 23486 23459 22528 96.03 95.92 95.98 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 174 Log-likelihood = -44526.168792 Norm (log-likelihood gradient vector) = 940.760965 Norm (lambda vector) = 264.136749 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4568 97.92 98.60 98.26 b-np 12220 12225 11989 98.07 98.11 98.09 i-np 13660 13721 13421 97.81 98.25 98.03 o 6349 6335 6186 97.65 97.43 97.54 b-vp 4768 4734 4637 97.95 97.25 97.60 i-vp 2602 2652 2528 95.32 97.16 96.23 b-adjp 384 371 322 86.79 83.85 85.30 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 820 726 88.54 88.32 88.43 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 491 459 93.48 91.25 92.35 i-adjp 152 136 115 84.56 75.66 79.86 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.59 74.54 76.51 Avg2. 46451 46451 45226 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12225 11772 96.29 96.33 96.31 pp 4633 4665 4567 97.90 98.58 98.24 vp 4768 4734 4551 96.13 95.45 95.79 sbar 503 491 459 93.48 91.25 92.35 adjp 384 371 317 85.44 82.55 83.97 advp 822 820 723 88.17 87.96 88.06 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.24 80.01 81.11 Avg2. 23486 23462 22533 96.04 95.94 95.99 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 175 Log-likelihood = -44293.152189 Norm (log-likelihood gradient vector) = 989.058129 Norm (lambda vector) = 264.388202 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4570 97.90 98.64 98.27 b-np 12220 12221 11985 98.07 98.08 98.07 i-np 13660 13730 13423 97.76 98.27 98.01 o 6349 6329 6181 97.66 97.35 97.51 b-vp 4768 4734 4636 97.93 97.23 97.58 i-vp 2602 2651 2527 95.32 97.12 96.21 b-adjp 384 369 321 86.99 83.59 85.26 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 821 726 88.43 88.32 88.37 i-advp 100 72 65 90.28 65.00 75.58 b-sbar 503 488 458 93.85 91.05 92.43 i-adjp 152 137 115 83.94 75.66 79.58 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.41 74.52 76.41 Avg2. 46451 46451 45217 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12221 11769 96.30 96.31 96.31 pp 4633 4668 4568 97.86 98.60 98.23 vp 4768 4734 4550 96.11 95.43 95.77 sbar 503 488 458 93.85 91.05 92.43 adjp 384 369 316 85.64 82.29 83.93 advp 822 821 723 88.06 87.96 88.01 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.28 79.96 81.10 Avg2. 23486 23457 22528 96.04 95.92 95.98 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 176 Log-likelihood = -43875.214832 Norm (log-likelihood gradient vector) = 1282.223801 Norm (lambda vector) = 265.199781 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4567 97.79 98.58 98.18 b-np 12220 12281 12009 97.79 98.27 98.03 i-np 13660 13591 13341 98.16 97.66 97.91 o 6349 6387 6203 97.12 97.70 97.41 b-vp 4768 4737 4636 97.87 97.23 97.55 i-vp 2602 2652 2527 95.29 97.12 96.19 b-adjp 384 376 325 86.44 84.64 85.53 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 824 725 87.99 88.20 88.09 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 493 457 92.70 90.85 91.77 i-adjp 152 140 117 83.57 76.97 80.14 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.25 74.20 76.17 Avg2. 46451 46451 45181 97.27 97.27 97.27 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12281 11781 95.93 96.41 96.17 pp 4633 4670 4565 97.75 98.53 98.14 vp 4768 4737 4551 96.07 95.45 95.76 sbar 503 493 457 92.70 90.85 91.77 adjp 384 376 320 85.11 83.33 84.21 advp 822 824 722 87.62 87.83 87.73 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.08 80.04 81.05 Avg2. 23486 23536 22540 95.77 95.97 95.87 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 177 Log-likelihood = -43672.797125 Norm (log-likelihood gradient vector) = 5514.268382 Norm (lambda vector) = 267.488675 Log-likelihood and gradient computational time: 321 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4569 97.84 98.62 98.23 b-np 12220 12250 12000 97.96 98.20 98.08 i-np 13660 13670 13395 97.99 98.06 98.02 o 6349 6356 6193 97.44 97.54 97.49 b-vp 4768 4734 4636 97.93 97.23 97.58 i-vp 2602 2649 2527 95.39 97.12 96.25 b-adjp 384 372 323 86.83 84.11 85.45 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 822 725 88.20 88.20 88.20 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 487 457 93.84 90.85 92.32 i-adjp 152 140 117 83.57 76.97 80.14 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.46 74.60 76.48 Avg2. 46451 46451 45217 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12250 11783 96.19 96.42 96.31 pp 4633 4670 4567 97.79 98.58 98.18 vp 4768 4734 4551 96.13 95.45 95.79 sbar 503 487 457 93.84 90.85 92.32 adjp 384 372 318 85.48 82.81 84.13 advp 822 822 722 87.83 87.83 87.83 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 79.99 81.13 Avg2. 23486 23490 22542 95.96 95.98 95.97 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 178 Log-likelihood = -43574.973103 Norm (log-likelihood gradient vector) = 2451.616551 Norm (lambda vector) = 266.171954 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4669 4568 97.84 98.60 98.22 b-np 12220 12240 11997 98.01 98.18 98.09 i-np 13660 13683 13402 97.95 98.11 98.03 o 6349 6348 6193 97.56 97.54 97.55 b-vp 4768 4736 4638 97.93 97.27 97.60 i-vp 2602 2648 2527 95.43 97.12 96.27 b-adjp 384 374 325 86.90 84.64 85.75 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 823 725 88.09 88.20 88.15 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 492 458 93.09 91.05 92.06 i-adjp 152 138 117 84.78 76.97 80.69 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.40 74.22 76.25 Avg2. 46451 46451 45224 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12240 11783 96.27 96.42 96.35 pp 4633 4669 4566 97.79 98.55 98.17 vp 4768 4736 4554 96.16 95.51 95.83 sbar 503 492 458 93.09 91.05 92.06 adjp 384 374 320 85.56 83.33 84.43 advp 822 823 722 87.73 87.83 87.78 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.22 80.07 81.13 Avg2. 23486 23489 22547 95.99 96.00 96.00 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 179 Log-likelihood = -43103.034829 Norm (log-likelihood gradient vector) = 1330.124267 Norm (lambda vector) = 267.309661 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4569 97.88 98.62 98.25 b-np 12220 12218 11987 98.11 98.09 98.10 i-np 13660 13724 13420 97.78 98.24 98.01 o 6349 6328 6185 97.74 97.42 97.58 b-vp 4768 4739 4641 97.93 97.34 97.63 i-vp 2602 2649 2529 95.47 97.19 96.32 b-adjp 384 373 324 86.86 84.38 85.60 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 824 728 88.35 88.56 88.46 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 493 459 93.10 91.25 92.17 i-adjp 152 136 116 85.29 76.32 80.56 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.55 74.20 76.31 Avg2. 46451 46451 45232 97.38 97.38 97.38 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12218 11774 96.37 96.35 96.36 pp 4633 4668 4567 97.84 98.58 98.20 vp 4768 4739 4557 96.16 95.57 95.87 sbar 503 493 459 93.10 91.25 92.17 adjp 384 373 320 85.79 83.33 84.54 advp 822 824 725 87.99 88.20 88.09 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 80.12 81.19 Avg2. 23486 23470 22546 96.06 96.00 96.03 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 180 Log-likelihood = -42703.032789 Norm (log-likelihood gradient vector) = 899.361543 Norm (lambda vector) = 268.468627 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4666 4567 97.88 98.58 98.23 b-np 12220 12212 11983 98.12 98.06 98.09 i-np 13660 13730 13424 97.77 98.27 98.02 o 6349 6319 6180 97.80 97.34 97.57 b-vp 4768 4745 4641 97.81 97.34 97.57 i-vp 2602 2653 2529 95.33 97.19 96.25 b-adjp 384 373 324 86.86 84.38 85.60 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 823 726 88.21 88.32 88.27 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 494 459 92.91 91.25 92.08 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.53 74.22 76.31 Avg2. 46451 46451 45224 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12212 11769 96.37 96.31 96.34 pp 4633 4666 4565 97.84 98.53 98.18 vp 4768 4745 4557 96.04 95.57 95.81 sbar 503 494 459 92.91 91.25 92.08 adjp 384 373 320 85.79 83.33 84.54 advp 822 823 723 87.85 87.96 87.90 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.25 80.09 81.15 Avg2. 23486 23468 22537 96.03 95.96 96.00 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 181 Log-likelihood = -42475.289082 Norm (log-likelihood gradient vector) = 1102.086845 Norm (lambda vector) = 269.171346 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4664 4565 97.88 98.53 98.20 b-np 12220 12210 11980 98.12 98.04 98.08 i-np 13660 13734 13424 97.74 98.27 98.01 o 6349 6321 6180 97.77 97.34 97.55 b-vp 4768 4745 4641 97.81 97.34 97.57 i-vp 2602 2651 2529 95.40 97.19 96.29 b-adjp 384 374 324 86.63 84.38 85.49 i-pp 52 40 34 85.00 65.38 73.91 b-advp 822 821 724 88.19 88.08 88.13 i-advp 100 71 65 91.55 65.00 76.02 b-sbar 503 495 459 92.73 91.25 91.98 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.50 74.20 76.29 Avg2. 46451 46451 45217 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12210 11766 96.36 96.28 96.32 pp 4633 4664 4563 97.83 98.49 98.16 vp 4768 4745 4558 96.06 95.60 95.83 sbar 503 495 459 92.73 91.25 91.98 adjp 384 374 320 85.56 83.33 84.43 advp 822 821 721 87.82 87.71 87.77 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.20 80.06 81.12 Avg2. 23486 23464 22531 96.02 95.93 95.98 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 182 Log-likelihood = -42061.974642 Norm (log-likelihood gradient vector) = 1220.777731 Norm (lambda vector) = 270.534017 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4663 4565 97.90 98.53 98.21 b-np 12220 12180 11957 98.17 97.85 98.01 i-np 13660 13800 13443 97.41 98.41 97.91 o 6349 6289 6156 97.89 96.96 97.42 b-vp 4768 4739 4639 97.89 97.29 97.59 i-vp 2602 2655 2529 95.25 97.19 96.21 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 820 724 88.29 88.08 88.19 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 496 458 92.34 91.05 91.69 i-adjp 152 139 117 84.17 76.97 80.41 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.67 75.02 76.80 Avg2. 46451 46451 45185 97.27 97.27 97.27 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12180 11725 96.26 95.95 96.11 pp 4633 4663 4563 97.86 98.49 98.17 vp 4768 4739 4553 96.08 95.49 95.78 sbar 503 496 458 92.34 91.05 91.69 adjp 384 370 317 85.68 82.55 84.08 advp 822 820 720 87.80 87.59 87.70 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.23 80.53 81.37 Avg2. 23486 23424 22481 95.97 95.72 95.85 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 183 Log-likelihood = -41717.699038 Norm (log-likelihood gradient vector) = 3476.936754 Norm (lambda vector) = 273.179260 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4660 4564 97.94 98.51 98.22 b-np 12220 12208 11977 98.11 98.01 98.06 i-np 13660 13744 13431 97.72 98.32 98.02 o 6349 6320 6177 97.74 97.29 97.51 b-vp 4768 4740 4641 97.91 97.34 97.62 i-vp 2602 2654 2529 95.29 97.19 96.23 b-adjp 384 369 320 86.72 83.33 84.99 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 821 723 88.06 87.96 88.01 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 497 459 92.35 91.25 91.80 i-adjp 152 138 117 84.78 76.97 80.69 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.70 75.03 76.82 Avg2. 46451 46451 45214 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12208 11759 96.32 96.23 96.27 pp 4633 4660 4562 97.90 98.47 98.18 vp 4768 4740 4556 96.12 95.55 95.84 sbar 503 497 459 92.35 91.25 91.80 adjp 384 369 316 85.64 82.29 83.93 advp 822 821 719 87.58 87.47 87.52 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.22 80.55 81.37 Avg2. 23486 23451 22516 96.01 95.87 95.94 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 184 Log-likelihood = -41275.509020 Norm (log-likelihood gradient vector) = 1251.027037 Norm (lambda vector) = 273.586351 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4661 4565 97.94 98.53 98.24 b-np 12220 12221 11983 98.05 98.06 98.06 i-np 13660 13729 13423 97.77 98.27 98.02 o 6349 6328 6180 97.66 97.34 97.50 b-vp 4768 4736 4639 97.95 97.29 97.62 i-vp 2602 2656 2529 95.22 97.19 96.20 b-adjp 384 368 320 86.96 83.33 85.11 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 821 723 88.06 87.96 88.01 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 494 457 92.51 90.85 91.68 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.74 75.01 76.83 Avg2. 46451 46451 45212 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12221 11765 96.27 96.28 96.27 pp 4633 4661 4563 97.90 98.49 98.19 vp 4768 4736 4553 96.14 95.49 95.81 sbar 503 494 457 92.51 90.85 91.68 adjp 384 368 316 85.87 82.29 84.04 advp 822 821 719 87.58 87.47 87.52 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.26 80.51 81.37 Avg2. 23486 23457 22518 96.00 95.88 95.94 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 185 Log-likelihood = -41123.088261 Norm (log-likelihood gradient vector) = 760.563684 Norm (lambda vector) = 273.800183 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4568 97.92 98.60 98.26 b-np 12220 12232 11985 97.98 98.08 98.03 i-np 13660 13715 13414 97.81 98.20 98.00 o 6349 6331 6178 97.58 97.31 97.44 b-vp 4768 4733 4638 97.99 97.27 97.63 i-vp 2602 2656 2529 95.22 97.19 96.20 b-adjp 384 369 321 86.99 83.59 85.26 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 823 724 87.97 88.08 88.02 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 491 456 92.87 90.66 91.75 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.52 74.18 76.29 Avg2. 46451 46451 45204 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11766 96.19 96.28 96.24 pp 4633 4665 4566 97.88 98.55 98.21 vp 4768 4733 4551 96.15 95.45 95.80 sbar 503 491 456 92.87 90.66 91.75 adjp 384 369 317 85.91 82.55 84.20 advp 822 823 720 87.48 87.59 87.54 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.21 79.90 81.04 Avg2. 23486 23468 22520 95.96 95.89 95.92 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 186 Log-likelihood = -40916.261486 Norm (log-likelihood gradient vector) = 1001.566861 Norm (lambda vector) = 274.531286 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4571 97.92 98.66 98.29 b-np 12220 12228 11983 98.00 98.06 98.03 i-np 13660 13715 13413 97.80 98.19 97.99 o 6349 6331 6178 97.58 97.31 97.44 b-vp 4768 4732 4641 98.08 97.34 97.71 i-vp 2602 2655 2530 95.29 97.23 96.25 b-adjp 384 371 323 87.06 84.11 85.56 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 824 726 88.11 88.32 88.21 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 489 456 93.25 90.66 91.94 i-adjp 152 138 117 84.78 76.97 80.69 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.46 74.22 76.28 Avg2. 46451 46451 45212 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11765 96.21 96.28 96.25 pp 4633 4668 4569 97.88 98.62 98.25 vp 4768 4732 4554 96.24 95.51 95.87 sbar 503 489 456 93.25 90.66 91.94 adjp 384 371 319 85.98 83.07 84.50 advp 822 824 722 87.62 87.83 87.73 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.28 79.99 81.12 Avg2. 23486 23467 22529 96.00 95.93 95.96 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 187 Log-likelihood = -40574.078046 Norm (log-likelihood gradient vector) = 1063.319654 Norm (lambda vector) = 275.667250 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4569 97.84 98.62 98.23 b-np 12220 12311 12012 97.57 98.30 97.93 i-np 13660 13539 13298 98.22 97.35 97.78 o 6349 6408 6206 96.85 97.75 97.30 b-vp 4768 4735 4640 97.99 97.32 97.65 i-vp 2602 2663 2533 95.12 97.35 96.22 b-adjp 384 370 322 87.03 83.85 85.41 i-pp 52 42 35 83.33 67.31 74.47 b-advp 822 820 725 88.41 88.20 88.31 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 494 458 92.71 91.05 91.88 i-adjp 152 140 118 84.29 77.63 80.82 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.30 74.25 76.22 Avg2. 46451 46451 45155 97.21 97.21 97.21 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12311 11772 95.62 96.33 95.98 pp 4633 4670 4566 97.77 98.55 98.16 vp 4768 4735 4554 96.18 95.51 95.84 sbar 503 494 458 92.71 91.05 91.88 adjp 384 370 319 86.22 83.07 84.62 advp 822 820 721 87.93 87.71 87.82 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.14 80.02 81.06 Avg2. 23486 23556 22534 95.66 95.95 95.80 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 188 Log-likelihood = -40998.224010 Norm (log-likelihood gradient vector) = 8037.793170 Norm (lambda vector) = 279.715781 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4572 97.90 98.68 98.29 b-np 12220 12251 11996 97.92 98.17 98.04 i-np 13660 13664 13386 97.97 97.99 97.98 o 6349 6352 6188 97.42 97.46 97.44 b-vp 4768 4733 4640 98.04 97.32 97.67 i-vp 2602 2663 2533 95.12 97.35 96.22 b-adjp 384 369 321 86.99 83.59 85.26 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 822 726 88.32 88.32 88.32 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 490 457 93.27 90.85 92.04 i-adjp 152 137 116 84.67 76.32 80.28 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.45 74.18 76.26 Avg2. 46451 46451 45209 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12251 11775 96.11 96.36 96.24 pp 4633 4670 4570 97.86 98.64 98.25 vp 4768 4733 4554 96.22 95.51 95.86 sbar 503 490 457 93.27 90.85 92.04 adjp 384 369 317 85.91 82.55 84.20 advp 822 822 722 87.83 87.83 87.83 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 79.97 81.11 Avg2. 23486 23490 22539 95.95 95.97 95.96 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 189 Log-likelihood = -40320.822373 Norm (log-likelihood gradient vector) = 3127.865668 Norm (lambda vector) = 277.246400 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4573 97.92 98.70 98.31 b-np 12220 12228 11985 98.01 98.08 98.04 i-np 13660 13706 13407 97.82 98.15 97.98 o 6349 6331 6177 97.57 97.29 97.43 b-vp 4768 4733 4639 98.01 97.29 97.65 i-vp 2602 2663 2533 95.12 97.35 96.22 b-adjp 384 369 322 87.26 83.85 85.52 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 821 726 88.43 88.32 88.37 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 492 459 93.29 91.25 92.26 i-adjp 152 137 116 84.67 76.32 80.28 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.57 74.63 76.55 Avg2. 46451 46451 45212 97.33 97.33 97.33 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12228 11764 96.21 96.27 96.24 pp 4633 4670 4571 97.88 98.66 98.27 vp 4768 4733 4553 96.20 95.49 95.84 sbar 503 492 459 93.29 91.25 92.26 adjp 384 369 318 86.18 82.81 84.46 advp 822 821 722 87.94 87.83 87.89 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.33 80.03 81.16 Avg2. 23486 23468 22531 96.01 95.93 95.97 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 190 Log-likelihood = -39874.103162 Norm (log-likelihood gradient vector) = 1147.559996 Norm (lambda vector) = 278.241434 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4574 97.92 98.73 98.32 b-np 12220 12219 11981 98.05 98.04 98.05 i-np 13660 13724 13417 97.76 98.22 97.99 o 6349 6322 6173 97.64 97.23 97.44 b-vp 4768 4735 4640 97.99 97.32 97.65 i-vp 2602 2664 2533 95.08 97.35 96.20 b-adjp 384 368 321 87.23 83.59 85.37 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 819 725 88.52 88.20 88.36 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 491 459 93.48 91.25 92.35 i-adjp 152 137 116 84.67 76.32 80.28 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.58 74.61 76.54 Avg2. 46451 46451 45214 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12219 11761 96.25 96.24 96.25 pp 4633 4671 4572 97.88 98.68 98.28 vp 4768 4735 4554 96.18 95.51 95.84 sbar 503 491 459 93.48 91.25 92.35 adjp 384 368 317 86.14 82.55 84.31 advp 822 819 721 88.03 87.71 87.87 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.36 79.99 81.16 Avg2. 23486 23458 22528 96.04 95.92 95.98 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 191 Log-likelihood = -39642.215079 Norm (log-likelihood gradient vector) = 857.212578 Norm (lambda vector) = 278.643715 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4572 97.94 98.68 98.31 b-np 12220 12214 11980 98.08 98.04 98.06 i-np 13660 13737 13427 97.74 98.29 98.02 o 6349 6317 6172 97.70 97.21 97.46 b-vp 4768 4737 4639 97.93 97.29 97.61 i-vp 2602 2665 2533 95.05 97.35 96.18 b-adjp 384 368 321 87.23 83.59 85.37 i-pp 52 41 35 85.37 67.31 75.27 b-advp 822 816 723 88.60 87.96 88.28 i-advp 100 71 64 90.14 64.00 74.85 b-sbar 503 492 459 93.29 91.25 92.26 i-adjp 152 136 116 85.29 76.32 80.56 b-prt 126 133 125 93.98 99.21 96.53 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 12 11 91.67 68.75 78.57 i-conjp 24 21 20 95.24 83.33 88.89 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.61 74.60 76.55 Avg2. 46451 46451 45217 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12214 11761 96.29 96.24 96.27 pp 4633 4668 4570 97.90 98.64 98.27 vp 4768 4737 4553 96.12 95.49 95.80 sbar 503 492 459 93.29 91.25 92.26 adjp 384 368 317 86.14 82.55 84.31 advp 822 816 719 88.11 87.47 87.79 prt 126 133 125 93.98 99.21 96.53 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 12 11 91.67 68.75 78.57 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.35 79.96 81.14 Avg2. 23486 23450 22523 96.05 95.90 95.97 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 192 Log-likelihood = -39394.402949 Norm (log-likelihood gradient vector) = 1041.708621 Norm (lambda vector) = 279.135502 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4571 97.92 98.66 98.29 b-np 12220 12211 11979 98.10 98.03 98.06 i-np 13660 13750 13436 97.72 98.36 98.04 o 6349 6313 6172 97.77 97.21 97.49 b-vp 4768 4736 4638 97.93 97.27 97.60 i-vp 2602 2664 2533 95.08 97.35 96.20 b-adjp 384 367 320 87.19 83.33 85.22 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 816 723 88.60 87.96 88.28 i-advp 100 70 64 91.43 64.00 75.29 b-sbar 503 490 458 93.47 91.05 92.25 i-adjp 152 135 116 85.93 76.32 80.84 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 13 10 76.92 83.33 80.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.81 75.00 76.86 Avg2. 46451 46451 45222 97.35 97.35 97.35 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12211 11761 96.31 96.24 96.28 pp 4633 4668 4569 97.88 98.62 98.25 vp 4768 4736 4553 96.14 95.49 95.81 sbar 503 490 458 93.47 91.05 92.25 adjp 384 367 317 86.38 82.55 84.42 advp 822 816 719 88.11 87.47 87.79 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.39 80.56 81.47 Avg2. 23486 23445 22522 96.06 95.90 95.98 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 193 Log-likelihood = -38886.675000 Norm (log-likelihood gradient vector) = 1235.848272 Norm (lambda vector) = 280.419337 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4570 97.86 98.64 98.25 b-np 12220 12233 11991 98.02 98.13 98.07 i-np 13660 13700 13413 97.91 98.19 98.05 o 6349 6339 6187 97.60 97.45 97.53 b-vp 4768 4743 4640 97.83 97.32 97.57 i-vp 2602 2652 2528 95.32 97.16 96.23 b-adjp 384 367 320 87.19 83.33 85.22 i-pp 52 39 34 87.18 65.38 74.73 b-advp 822 819 724 88.40 88.08 88.24 i-advp 100 75 67 89.33 67.00 76.57 b-sbar 503 487 456 93.63 90.66 92.12 i-adjp 152 134 117 87.31 76.97 81.82 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 14 13 92.86 81.25 86.67 i-conjp 24 23 22 95.65 91.67 93.62 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.72 75.27 76.96 Avg2. 46451 46451 45226 97.36 97.36 97.36 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12233 11781 96.31 96.41 96.36 pp 4633 4670 4568 97.82 98.60 98.20 vp 4768 4743 4555 96.04 95.53 95.78 sbar 503 487 456 93.63 90.66 92.12 adjp 384 367 317 86.38 82.55 84.42 advp 822 819 721 88.03 87.71 87.87 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 14 13 92.86 81.25 86.67 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.43 81.19 81.81 Avg2. 23486 23477 22544 96.03 95.99 96.01 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 194 Log-likelihood = -38341.031061 Norm (log-likelihood gradient vector) = 2854.469527 Norm (lambda vector) = 283.362663 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4569 97.77 98.62 98.19 b-np 12220 12220 11982 98.05 98.05 98.05 i-np 13660 13731 13426 97.78 98.29 98.03 o 6349 6324 6179 97.71 97.32 97.51 b-vp 4768 4742 4640 97.85 97.32 97.58 i-vp 2602 2651 2528 95.36 97.16 96.25 b-adjp 384 367 320 87.19 83.33 85.22 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 817 722 88.37 87.83 88.10 i-advp 100 74 67 90.54 67.00 77.01 b-sbar 503 486 455 93.62 90.46 92.01 i-adjp 152 134 117 87.31 76.97 81.82 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.53 74.72 76.58 Avg2. 46451 46451 45216 97.34 97.34 97.34 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12220 11764 96.27 96.27 96.27 pp 4633 4673 4567 97.73 98.58 98.15 vp 4768 4742 4554 96.04 95.51 95.77 sbar 503 486 455 93.62 90.46 92.01 adjp 384 367 317 86.38 82.55 84.42 advp 822 817 719 88.00 87.47 87.74 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.36 80.50 81.42 Avg2. 23486 23462 22521 95.99 95.89 95.94 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 195 Log-likelihood = -37818.173803 Norm (log-likelihood gradient vector) = 901.612877 Norm (lambda vector) = 284.167196 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4673 4569 97.77 98.62 98.19 b-np 12220 12220 11981 98.04 98.04 98.04 i-np 13660 13732 13421 97.74 98.25 97.99 o 6349 6322 6176 97.69 97.28 97.48 b-vp 4768 4740 4638 97.85 97.27 97.56 i-vp 2602 2653 2528 95.29 97.16 96.21 b-adjp 384 367 320 87.19 83.33 85.22 i-pp 52 41 34 82.93 65.38 73.12 b-advp 822 817 722 88.37 87.83 88.10 i-advp 100 74 67 90.54 67.00 77.01 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 134 117 87.31 76.97 81.82 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.51 74.72 76.57 Avg2. 46451 46451 45205 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12220 11759 96.23 96.23 96.23 pp 4633 4673 4567 97.73 98.58 98.15 vp 4768 4740 4551 96.01 95.45 95.73 sbar 503 487 455 93.43 90.46 91.92 adjp 384 367 317 86.38 82.55 84.42 advp 822 817 719 88.00 87.47 87.74 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.34 80.49 81.41 Avg2. 23486 23461 22513 95.96 95.86 95.91 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 196 Log-likelihood = -37723.988099 Norm (log-likelihood gradient vector) = 746.453513 Norm (lambda vector) = 284.308411 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4568 97.79 98.60 98.19 b-np 12220 12229 11986 98.01 98.09 98.05 i-np 13660 13718 13414 97.78 98.20 97.99 o 6349 6325 6178 97.68 97.31 97.49 b-vp 4768 4737 4637 97.89 97.25 97.57 i-vp 2602 2653 2528 95.29 97.16 96.21 b-adjp 384 367 321 87.47 83.59 85.49 i-pp 52 42 34 80.95 65.38 72.34 b-advp 822 818 724 88.51 88.08 88.29 i-advp 100 75 67 89.33 67.00 76.57 b-sbar 503 489 457 93.46 90.85 92.14 i-adjp 152 136 117 86.03 76.97 81.25 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.31 74.76 76.49 Avg2. 46451 46451 45208 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12229 11763 96.19 96.26 96.22 pp 4633 4671 4566 97.75 98.55 98.15 vp 4768 4737 4549 96.03 95.41 95.72 sbar 503 489 457 93.46 90.85 92.14 adjp 384 367 318 86.65 82.81 84.69 advp 822 818 721 88.14 87.71 87.93 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.38 80.58 81.47 Avg2. 23486 23468 22519 95.96 95.88 95.92 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 197 Log-likelihood = -37434.046207 Norm (log-likelihood gradient vector) = 917.759105 Norm (lambda vector) = 285.582705 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4671 4566 97.75 98.55 98.15 b-np 12220 12208 11971 98.06 97.96 98.01 i-np 13660 13765 13422 97.51 98.26 97.88 o 6349 6303 6162 97.76 97.05 97.41 b-vp 4768 4737 4638 97.91 97.27 97.59 i-vp 2602 2651 2528 95.36 97.16 96.25 b-adjp 384 368 320 86.96 83.33 85.11 i-pp 52 42 34 80.95 65.38 72.34 b-advp 822 818 723 88.39 87.96 88.17 i-advp 100 75 66 88.00 66.00 75.43 b-sbar 503 487 455 93.43 90.46 91.92 i-adjp 152 135 116 85.93 76.32 80.84 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.20 74.62 76.37 Avg2. 46451 46451 45178 97.26 97.26 97.26 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12208 11747 96.22 96.13 96.18 pp 4633 4671 4564 97.71 98.51 98.11 vp 4768 4737 4550 96.05 95.43 95.74 sbar 503 487 455 93.43 90.46 91.92 adjp 384 368 317 86.14 82.55 84.31 advp 822 818 718 87.78 87.35 87.56 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 80.46 81.37 Avg2. 23486 23446 22496 95.95 95.78 95.87 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds Iteration: 198 Log-likelihood = -37188.310332 Norm (log-likelihood gradient vector) = 2301.188432 Norm (lambda vector) = 287.682988 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 321 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4670 4565 97.75 98.53 98.14 b-np 12220 12234 11982 97.94 98.05 98.00 i-np 13660 13710 13408 97.80 98.16 97.98 o 6349 6327 6177 97.63 97.29 97.46 b-vp 4768 4735 4637 97.93 97.25 97.59 i-vp 2602 2652 2530 95.40 97.23 96.31 b-adjp 384 370 321 86.76 83.59 85.15 i-pp 52 43 34 79.07 65.38 71.58 b-advp 822 820 723 88.17 87.96 88.06 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 488 456 93.44 90.66 92.03 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.13 74.69 76.37 Avg2. 46451 46451 45193 97.29 97.29 97.29 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12234 11762 96.14 96.25 96.20 pp 4633 4670 4563 97.71 98.49 98.10 vp 4768 4735 4550 96.09 95.43 95.76 sbar 503 488 456 93.44 90.66 92.03 adjp 384 370 318 85.95 82.81 84.35 advp 822 820 718 87.56 87.35 87.45 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.25 80.52 81.37 Avg2. 23486 23474 22512 95.90 95.85 95.88 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 199 Log-likelihood = -36925.837487 Norm (log-likelihood gradient vector) = 1377.739628 Norm (lambda vector) = 288.775525 Log-likelihood and gradient computational time: 320 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4668 4565 97.79 98.53 98.16 b-np 12220 12232 11986 97.99 98.09 98.04 i-np 13660 13712 13413 97.82 98.19 98.01 o 6349 6328 6181 97.68 97.35 97.52 b-vp 4768 4735 4637 97.93 97.25 97.59 i-vp 2602 2655 2530 95.29 97.23 96.25 b-adjp 384 369 321 86.99 83.59 85.26 i-pp 52 43 34 79.07 65.38 71.58 b-advp 822 818 722 88.26 87.83 88.05 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 489 457 93.46 90.85 92.14 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 13 12 92.31 75.00 82.76 i-conjp 24 22 21 95.45 87.50 91.30 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 78.15 74.70 76.38 Avg2. 46451 46451 45206 97.32 97.32 97.32 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11768 96.21 96.30 96.25 pp 4633 4668 4563 97.75 98.49 98.12 vp 4768 4735 4550 96.09 95.43 95.76 sbar 503 489 457 93.46 90.85 92.14 adjp 384 369 318 86.18 82.81 84.46 advp 822 818 717 87.65 87.23 87.44 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 13 12 92.31 75.00 82.76 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 82.29 80.53 81.40 Avg2. 23486 23468 22518 95.95 95.88 95.92 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 355 seconds Iteration: 200 Log-likelihood = -36688.950623 Norm (log-likelihood gradient vector) = 875.272330 Norm (lambda vector) = 289.502541 Log-likelihood and gradient computational time: 319 seconds Training iteration elapsed: 320 seconds Label-based performance evaluation: Label Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- b-pp 4633 4665 4563 97.81 98.49 98.15 b-np 12220 12232 11985 97.98 98.08 98.03 i-np 13660 13707 13411 97.84 98.18 98.01 o 6349 6333 6185 97.66 97.42 97.54 b-vp 4768 4737 4636 97.87 97.23 97.55 i-vp 2602 2656 2530 95.26 97.23 96.23 b-adjp 384 369 321 86.99 83.59 85.26 i-pp 52 41 33 80.49 63.46 70.97 b-advp 822 816 720 88.24 87.59 87.91 i-advp 100 74 66 89.19 66.00 75.86 b-sbar 503 489 457 93.46 90.85 92.14 i-adjp 152 137 117 85.40 76.97 80.97 b-prt 126 134 125 93.28 99.21 96.15 i-sbar 12 12 9 75.00 75.00 75.00 b-conjp 16 15 13 86.67 81.25 83.87 i-conjp 24 24 22 91.67 91.67 91.67 b-intj 4 0 0 0.00 0.00 0.00 b-lst 10 10 10 100.00 100.00 100.00 b-ucp 0 0 0 0.00 0.00 0.00 i-ucp 0 0 0 0.00 0.00 0.00 i-intj 12 0 0 0.00 0.00 0.00 i-prt 0 0 0 0.00 0.00 0.00 i-lst 2 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 77.74 75.11 76.40 Avg2. 46451 46451 45203 97.31 97.31 97.31 Chunk-based performance evaluation: Chunk Manual Model Match Pre.(%) Rec.(%) F1-Measure(%) ----- ------ ----- ----- ------- ------- ------------- np 12220 12232 11766 96.19 96.28 96.24 pp 4633 4665 4561 97.77 98.45 98.11 vp 4768 4737 4549 96.03 95.41 95.72 sbar 503 489 457 93.46 90.85 92.14 adjp 384 369 318 86.18 82.81 84.46 advp 822 816 715 87.62 86.98 87.30 prt 126 134 125 93.28 99.21 96.15 lst 10 10 8 80.00 80.00 80.00 intj 4 0 0 0.00 0.00 0.00 conjp 16 15 13 86.67 81.25 83.87 ucp 0 0 0 0.00 0.00 0.00 ----- ------ ----- ----- ------- ------- ------------- Avg1. 81.72 81.12 81.42 Avg2. 23486 23467 22512 95.93 95.85 95.89 Current max chunk-based F1: 96.10 (iteration 168) Training iteration elapsed (including evaluation time): 354 seconds The training process elapsed: 70935 seconds