[NER] Averaged Perceptron, pre integration, 2016 10 18

Iteration: 10
Data Set: CoNLL2003 English(ner_default.config)

Log:

2016-10-18 08:58:39,913 MAIN [INFO]: Reading configurations from file: config/ner_default.config
2016-10-18 08:58:39,915 NER [INFO]: Initialising NER Tagger
2016-10-18 08:58:39,915 NER [INFO]: Tag File selected: file://ner_tagset.txt
2016-10-18 08:58:39,915 NER [INFO]: Initialisation Complete
2016-10-18 08:58:39,915 NER [INFO]: Loading training data
2016-10-18 09:08:01,523 NER [INFO]: Total sentence Number: 14987
2016-10-18 09:08:01,530 NER [INFO]: Training data loaded
2016-10-18 09:08:01,530 NER [INFO]: Starting Training Process
2016-10-18 11:55:17,394 NER [INFO]: Total Training Time(seconds): 10035.864167
2016-10-18 11:55:18,500 WEIGHTVECTOR [INFO]: Dumping Weight Vector to file://NER_Iter_10.db
2016-10-18 11:55:18,501 WEIGHTVECTOR [INFO]: Total Feature Num: 248418
2016-10-18 11:55:18,526 NER [INFO]: Starting evaluation process
2016-10-18 11:55:18,527 NER [INFO]: Loading evaluation data
2016-10-18 11:57:24,520 NER [INFO]: Total sentence Number: 3467
2016-10-18 11:57:24,523 NER [INFO]: Evaluation data loaded
2016-10-18 12:01:04,197 NER [INFO]: Total evaluation Time(seconds): 219.674527

conlleval.pl:

processed 51578 tokens with 5942 phrases; found: 5810 phrases; correct: 5017.
num_NER_tags: 1376.00%; correct_tags: 50202.00%; tokens: 51578.00%; accuracy: 97.33%; precision: 86.35%; recall: 84.43%; FB1: 85.38
LOC: precision: 89.99%; recall: 85.68%; FB1: 87.79 1749
MISC: precision: 85.54%; recall: 82.75%; FB1: 84.12 892
ORG: precision: 79.24%; recall: 78.00%; FB1: 78.62 1320
PER: precision: 88.37%; recall: 88.71%; FB1: 88.54 1849