A Novel Approach to Dropped Pronoun Translation

Proceedings of NAACL-HLT 2016, pages 983–993, San Diego, California, June 12-17, 2016. c⃝2016 Association for Computational Linguistics A Novel Approach to Dropped Pronoun Translation Longyue Wang† Zhaopeng Tu‡ Xiaojun Zhang† Hang Li‡ Andy Way† Qun Liu† †ADAPT Centre, School of Computing, Dublin City University, Ireland {lwang, xzhang, away, qliu}@computing.dcu.ie ‡Noah’s Ark Lab, Huawei Technologies, China {tu.zhaopeng, hangli.hl}@huawei.com Abstract Dropped Pronouns (DP) in which pronouns are frequently dropped in the source language but should be retained in the target language are challenge in machine translation. In re- sponse to this problem, we propose a semi- supervised approach to recall possibly missing pronouns in the translation. Firstly, we build training data for DP generation in which the DPs are automatically labelled according to the alignment information from a parallel cor- pus. Secondly, we build a deep learning-based DP generator for input sentences in decoding when no corresponding references exist. More specifically, the generation is two-phase: (1) DP position detection, which is modeled as a sequential labelling task with recurrent neural networks; and (2) DP prediction, which em- ploys a multilayer perceptron with rich fea- tures. Finally, we integrate the above outputs into our translation system to recall missing pronouns by both extracting rules from the DP-labelled training data and translating the DP-generated input sentences. Experimental results show that our approach achieves a sig- nificant improvement of 1.58 BLEU points in translation performance with 66% F-score for DP generation accuracy. 1 Introduction In pro-drop languages, certain classes of pronouns can be omitted to make the sentence compact yet comprehensible when the identity of the pronouns can be inferred from the context (Yang et al., 2015). Figure 1 shows an example, in which Chinese is a pro-drop language (Huang, 1984), while English is Figure 1: Examples of dropped pronouns in a parallel dialogue corpus. The Chinese pronouns in brackets are dropped. not (Haspelmath, 2001). On the Chinese side, the subject pronouns {你 (you), 我 (I)} and the object pronouns {它 (it), 你 (you)} are omitted in the di- alogue between Speakers A and B. These omis- sions may not be problems for humans since peo- ple can easily recall the missing pronouns from the context. However, this poses difficulties for Sta- tistical Machine Translation (SMT) from pro-drop languages (e.g. Chinese) to non-pro-drop languages (e.g. English), since translation of such missing pro- nouns cannot be normally reproduced. Generally, this phenomenon is more common in informal gen- res such as dialogues and conversations than oth- ers (Yang et al., 2015). We also validated this finding by analysing a large Chinese–English dialogue cor- pus which consists of 1M sentence pairs extracted from movie and TV episode subtitles. We found that there are 6.5M Chinese pronouns and 9.4M English pronouns, which shows that more than 2.9 million Chinese pronouns are missing. In response to this problem, we propose to find a general and replicable way to improve translation quality. The main challenge of this research is that training data for DP generation are scarce. Most 983 works either …

Content truncated to 500 words