Improving Statistical Machine Translation with a Multilingual Paraphrase Database

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1379–1390, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. Improving Statistical Machine Translation with a Multilingual Paraphrase Database Ramtin Mehdizadeh Seraj, Maryam Siahbani, Anoop Sarkar School of Computing Science Simon Fraser University Burnaby BC. Canada rmehdiza,msiahban,anoop@cs.sfu.ca Abstract The multilingual Paraphrase Database (PPDB) is a freely available automatically created resource of paraphrases in mul- tiple languages. In statistical machine translation, paraphrases can be used to provide translation for out-of-vocabulary (OOV) phrases. In this paper, we show that a graph propagation approach that uses PPDB paraphrases can be used to im- prove overall translation quality. We pro- vide an extensive comparison with previ- ous work and show that our PPDB-based method improves the BLEU score by up to 1.79 percent points. We show that our approach improves on the state of the art in three different settings: when faced with limited amount of parallel training data; a domain shift between training and test data; and handling a morpho- logically complex source language. Our PPDB-based method outperforms the use of distributional profiles from monolin- gual source data. 1 Introduction Translation coverage is a major concern in statis- tical machine translation (SMT) which relies on large amounts of parallel, sentence-aligned text. In (Callison-Burch et al., 2006), even with a training data size of 10 million word tokens, source vocab- ulary coverage in unseen data does not go above 90%. The problem is worse with multi-word OOV phrases. Copying OOVs to the output is the most common solution. However, even noisy transla- tions of OOVs can improve reordering and lan- guage model scores (Zhang et al., 2012). Translit- eration is useful but not a panacea for the OOV problem (Irvine and Callison-Burch, 2014b). We find and remove the named entities, dates, etc. in the source and focus on the use of paraphrases to help translate the remaining OOVs. In Sec. 5.2 we show that handling such OOVs correctly does im- prove translation scores. In this paper, we build on the following re- search: Bilingual lexicon induction is the task of learning translations of words from monolin- gual data in source and target languages (Schafer and Yarowsky, 2002; Koehn and Knight, 2002; Haghighi et al., 2008). The distributional pro- file (DP) approach uses context vectors to link words as potential paraphrases to translation can- didates (Rapp, 1995; Koehn and Knight, 2002; Haghighi et al., 2008; Garera et al., 2009). DPs have been used in SMT to assign translation can- didates to OOVs (Marton et al., 2009; Daum´e and Jagarlamudi, 2011; Irvine et al., 2013; Irvine and Callison-Burch, 2014a). Graph-based semi- supervised methods extend this approach and propagate translation candidates across a graph with phrasal nodes connected via weighted para- phrase relationships (Razmara et al., 2013; Saluja et al., 2014; Zhao et al., 2015). Saluja et al. (2014) extend paraphrases for SMT from the words to phrases, which we also do in this work. Bilin- gual pivoting uses parallel data instead of …

Content truncated to 500 words