Consistent Translation of Repeated Nouns using Syntactic and Semantic Cues

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 948–957, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Consistent Translation of Repeated Nouns using Syntactic and Semantic Cues Xiao Pu1,2, Laura Mascarell3 and Andrei Popescu-Belis1 1Idiap Research Institute, 1920 Martigny, CH 2 ´Ecole Polytechnique F´ed´erale de Lausanne, 1015 Lausanne, CH 3Institute of Computational Linguistics, University of Z¨urich, 8050 Z¨urich, CH {xiao.pu, andrei.popescu-belis}@idiap.ch {mascarell}@cl.uzh.ch Abstract We propose a method to decide whether two occurrences of the same noun in a source text should be translated consis- tently, i.e. using the same noun in the tar- get text as well. We train and test clas- sifiers that predict consistent translations based on lexical, syntactic, and seman- tic features. We first evaluate the ac- curacy of our classifiers intrinsically, in terms of the accuracy of consistency pre- dictions, over a subset of the UN Corpus. Then, we also evaluate them in combina- tion with phrase-based statistical MT sys- tems for Chinese-to-English and German- to-English. We compare the automatic post-editing of noun translations with the re-ranking of the translation hypotheses based on the classifiers’ output, and also use these methods in combination. This improves over the baseline and closes up to 50% of the gap in BLEU scores between the baseline and an oracle classifier. 1 Introduction The repetition of a noun in a text may be due to co-reference, i.e. repeated mentions of the same entity, or to mentions of two entities of the same type. But in other cases, two occurrences of the same noun may simply convey different meanings. The translation of repeated nouns depends, among other things, on the conveyed meanings: in case of co-reference or identical senses, they should likely be translated with the same word, while otherwise they should be translated with different words, if the target language distinguishes the two mean- ings. State-of-the-art machine translation systems do not address this challenge systematically, and translate two occurrences of the same noun inde- pendently, thus potentially introducing unwanted variations in translation. We exemplify this issue in Figure 1 for Chinese- to-English and German-to-English translations, with examples of inconsistent translations of a re- peated source noun by a baseline SMT system, as opposed to consistent translations in the reference. In Example 1, the system’s translation of the sec- ond occurrence of politik is mistaken and should be replaced by the first one (policy, not politics). In Example 2, although the first translation differs from the reference, it could be acceptable as a le- gitimate variation, although the second one (iden- tity documents) is more idiomatic and more fre- quent. Of course, in addition to these two exam- ples, there are other configurations of the six nouns involved in a consistency relation across source, candidate and reference translations, but they will be discussed below when designing the training and test data for our problem. In this paper, we aim to improve the transla- tion of repeated nouns by …

Content truncated to 500 words