Getting the Structure Right for Word Alignment: LEAF

Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pp. 51–60, Prague, June 2007. c⃝2007 Association for Computational Linguistics Getting the structure right for word alignment: LEAF Alexander Fraser ISI / University of Southern California 4676 Admiralty Way, Suite 1001 Marina del Rey, CA 90292 fraser@isi.edu Daniel Marcu ISI / University of Southern California 4676 Admiralty Way, Suite 1001 Marina del Rey, CA 90292 marcu@isi.edu Abstract Word alignment is the problem of annotating parallel text with translational correspon- dence. Previous generative word alignment models have made structural assumptions such as the 1-to-1, 1-to-N, or phrase-based consecutive word assumptions, while previ- ous discriminative models have either made such an assumption directly or used features derived from a generative model making one of these assumptions. We present a new gen- erative alignment model which avoids these structural limitations, and show that it is effective when trained using both unsuper- vised and semi-supervised training methods. 1 Introduction Several generative models and a large number of discriminatively trained models have been proposed in the literature to solve the problem of automatic word alignment of bitexts. The generative propos- als have required unrealistic assumptions about the structure of the word alignments. Two assumptions are particularly common. The first is the 1-to-N as- sumption, meaning that each source word generates zero or more target words, which requires heuristic techniques in order to obtain alignments suitable for training a SMT system. The second is the consec- utive word-based “phrasal SMT” assumption. This does not allow gaps, which can be used to particular advantage by SMT models which model hierarchi- cal structure. Previous discriminative models have either made such assumptions directly or used fea- tures from a generative model making such an as- sumption. Our objective is to automatically produce alignments which can be used to build high quality machine translation systems. These are presumably close to the alignments that trained bilingual speak- ers produce. Human annotated alignments often contain M-to-N alignments, where several source words are aligned to several target words and the re- sulting unit can not be further decomposed. Source or target words in a single unit are sometimes non- consecutive. In this paper, we describe a new generative model which directly models M-to-N non-consecutive word alignments. The rest of the paper is organized as follows. The generative story is presented, fol- lowed by the mathematical formulation. Details of the unsupervised training procedure are described. The generative model is then decomposed into fea- ture functions used in a log-linear model which is trained using a semi-supervised algorithm. Experi- ments show improvements in word alignment accu- racy and usage of the generated alignments in hier- archical and phrasal SMT systems results in an in- creased BLEU score. Previous work is discussed and this is followed by the conclusion. 2 LEAF: a generative word alignment model 2.1 Generative story We introduce a new generative story which enables the capture of non-consecutive M-to-N alignment structure. We have attempted to …

Content truncated to 500 words