Active Learning for Word Sense Disambiguation with Methods for Addressing the Class Imbalance Problem Export

@inproceedings{citeulike:1652188, abstract = {In this paper, we analyze the effect of resampling techniques, including undersampling and over-sampling used in active learning for word sense disambiguation (WSD). Experimental results show that under- sampling causes negative effects on active learning, but over-sampling is a relatively good choice. To alleviate the withinclass imbalance problem of over-sampling, we propose a bootstrap-based oversampling (BootOS) method that works better than ordinary over-sampling in active learning for WSD. Finally, we investigate when to stop active learning, and adopt two strategies, max-confidence and min-error, as stopping conditions for active learning. According to experimental results, we suggest a prediction solution by considering max-confidence as the upper bound and min-error as the lower bound for stopping conditions.}, author = {Zhu, Jingbo and Hovy, Eduard}, citeulike-article-id = {1652188}, citeulike-linkout-0 = {http://www.aclweb.org/anthology-new/D/D07/D07-1082.pdf}, journal = {Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning}, keywords = {active-learning, word-sense-disambiguation}, month = {June}, posted-at = {2007-09-13 16:25:54}, priority = {1}, title = {Active Learning for Word Sense Disambiguation with Methods for Addressing the Class Imbalance Problem}, url = {http://www.aclweb.org/anthology-new/D/D07/D07-1082.pdf}, year = {2007} }

TY - CONF ID - citeulike:1652188 L3 - citeulike-article-id:1652188 N2 - In this paper, we analyze the effect of resampling techniques, including undersampling and over-sampling used in active learning for word sense disambiguation (WSD). Experimental results show that under- sampling causes negative effects on active learning, but over-sampling is a relatively good choice. To alleviate the withinclass imbalance problem of over-sampling, we propose a bootstrap-based oversampling (BootOS) method that works better than ordinary over-sampling in active learning for WSD. Finally, we investigate when to stop active learning, and adopt two strategies, max-confidence and min-error, as stopping conditions for active learning. According to experimental results, we suggest a prediction solution by considering max-confidence as the upper bound and min-error as the lower bound for stopping conditions. JF - Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning TI - Active Learning for Word Sense Disambiguation with Methods for Addressing the Class Imbalance Problem KW - active-learning KW - word-sense-disambiguation AU - Zhu, Jingbo AU - Hovy, Eduard PY - 2007/06// UR - http://www.aclweb.org/anthology-new/D/D07/D07-1082.pdf ER -