Improved chunk-level reordering for statistical machine translation

Improved Chunk-level Reordering for Statistical Machine Translation Yuqi Zhang, Richard Zens and Hermann Ney Human Language Technology and Pattern Recognition Lehrstuhl f¨ur Informatik 6 Computer Science Department RWTH Aachen University, Germany Abstract Inspired by previous chunk-level reordering ap- proaches to statistical machine translation, this pa- per presents two methods to improve the reorder- ing at the chunk level. By introducing a new lat- tice weighting factor and by reordering the train- ing source data, an improvement is reported on TER and BLEU. Compared to the previous chunk- level reordering approach, the BLEU score im- proves 1.4% absolutely. The translation results are reported on IWSLT Chinese-English task. 1. Introduction In machine translation, reordering is one of the ma- jor problems, since different languages have dif- ferent word order requirements. In current phrase- based Statistical Machine Translation (SMT) sys- tems, distance-based reordering constraints are widely used, such as IBM constraints [1], local constraints [2] and distortion limit [3]. With these models phrase-based SMT is powerful in word re- ordering within short distance. However, long- distance reordering is still problematic. In order to solve the long-distance reordering problem, it has been realized that syntactic infor- mation should be used. Some approaches have ap- plied at the word-level, such as morphology [4], POS tags [5] and word classes [6]. They are particularly useful for the language with rich mor- phology for reducing the data sparseness. Another kinds of syntax reordering methods require parse trees, such as the work in [7], [8], [9], [10]. The parse tree is more powerful to capture the sentence structures. However, it is expensive to create tree structures and building a good quality parser is also a hard task. What we are interested in here is to use an in- termediate syntax between POS tag and parse tree: chunks, as the basic unit for reordering. It is not only because chunks are with more syntax than POS tags, but also they are closer to the defini- tion of a “phrase” in phrase-based SMT and easy to use. We have not found much work to do reorder- ing at the chunk level. Schafer [11] has devel- oped a word-chunk two levels syntactic transduc- tion which uses chunks on both language sides. It is a whole translation system. Here, we only ap- ply chunks on source language and are more in- terested in using chunk knowledge in the phrase- based translation framework. In this paper, we will improve the approach de- scribed in [12] by adding a weight model using the rules probability and repeating training on the reordered sentence pairs. In Section 3, the baseline systems are introduced. Section 4 is the main part of the paper, where the new methods to improve the baseline model are presented. Section 5 de- scribes the experiments and the analysis. Finally, Section 6 is the conclusion. 2. Related work In the previous chunk level reordering work, [12] has represented the reorderings generated with some rules in a weighted lattice. The lattice is weighted with …

Content truncated to 500 words