Fast Domain Adaptation for Part of Speech Tagging for Dialogues
Sandra Kübler, Eric Baucom · Recent Advances in Natural Language Processing · 2011
Part of speech tagging accuracy deteriorates severely when a tagger is used out of domain. We investigate a fast method for domain adaptation, which provides additional in-domain training data from an unannotated data set by applying POS taggers with different biases to the unannotated data set and then choosing the set of sentences on which the taggers agree. We show that we improve the accuracy of a trigram tagger, TnT, from 85.77% to 86.10%. In order to improve performance on unknown words, we investigate using active learning for learning ambiguity classes of domain specific words, yielding an accuracy of 89.15% for TnT.