tailieunhanh - Báo cáo khoa học: "Maximum Expected BLEU Training of Phrase and Lexicon Translation Models"

This paper proposes a new discriminative training method in constructing phrase and lexicon translation models. In order to reliably learn a myriad of parameters in these models, we propose an expected BLEU score-based utility function with KL regularization as the objective, and train the models on a large parallel dataset. | Maximum Expected BLEU Training of Phrase and Lexicon Translation Models Xiaodong He Microsoft Research Li Deng Microsoft Research One Microsoft Way Redmond WA USA xiaohe@ One Microsoft Way Redmond WA USA deng@ Abstract This paper proposes a new discriminative training method in constructing phrase and lexicon translation models. In order to reliably learn a myriad of parameters in these models we propose an expected BLEU score-based utility function with KL regularization as the objective and train the models on a large parallel dataset. For training we derive growth transformations for phrase and lexicon translation probabilities to iteratively improve the objective. The proposed method evaluated on the Europarl German-to-English dataset leads to a BLEU point improvement over a state-of-the-art baseline translation system. In IWSLT 2011 Benchmark our system using the proposed method achieves the best Chinese-to-English translation result on the task of translating TED talks. 1. Introduction Discriminative training is an active area in statistical machine translation SMT . Och et al. 2002 2003 Liang et al. 2006 Blunsom et al. 2008 Chiang et al. 2009 Foster et al 2010 Xiao et al. 2011 . Och 2003 proposed using a log-linear model to incorporate multiple features for translation and proposed a minimum error rate training MERT method to train the feature weights to optimize a desirable translation metric. While the log-linear model itself is discriminative the phrase and lexicon translation features which are among the most important components of SMT are derived from either generative models or heuristics Koehn et al. 2003 Brown et al. 1993 . Moreover the 292 parameters in the phrase and lexicon translation models are estimated by relative frequency or maximizing joint likelihood which may not correspond closely to the translation measure . bilingual evaluation understudy BLEU Papineni et al. 2002 . Therefore it is desirable to .