Proceedings of the Conference on Empirical Methods in Natural Language Processing: 17-18 May 1996, University of Pennsylvania, Philadelphia, Pa. USAEric Brill |
Contents
A Geometric Approach to Mapping Bitext Correspondence | 1 |
Methods and Measures | 23 |
Automatic Extraction of New Words from Japanese Texts using Generalized ForwardBackward | 48 |
5 other sections not shown
Common terms and phrases
accuracy AGR 3SG alignment ambiguity annotation approach baseline Basili bigram bigram model bitext map bitext space Bod's case-based learning chain chart parsing clause nounph collision sets Computational Linguistics constituent constraints context corpora delete derived DOP model dysfluencies error esl's estimate evaluation example expected word experiments feature set selection figure of merit frequency given grammar hand-crafted heuristic hypotheses language models learning algorithm lexical linguistic bias Machine Learning memory bias methods model merging model performance N-gram Natural Language Processing nonterminal nounph modph nounph noun nc output parameters parse tree parser part-of-speech part-of-speech tagging PCFG perplexity phrase probabilistic probability problem Proceedings recall recency bias relative pronoun representation rules semantic sentence sequence SIMR SIMR's speech speech recognition statistical STSG subtree syntactic Table test set tion token training corpus training data training set Treebank trigram unigram unknown word vadv variables verb verbph vn word segmentation word unigram