System Combination for Machine Translation through Paraphrasing

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1053–1058, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. System Combination for Machine Translation through Paraphrasing Wei-Yun Ma Institute of Information science Academia Sinica Taipei 115, Taiwan ma@iis.sinica.edu.tw Kathleen McKeown Department of Computer Science Columbia University New York, NY 10027, USA kathy@cs.columbia.edu Abstract In this paper, we propose a paraphrasing model to address the task of system com- bination for machine translation. We dy- namically learn hierarchical paraphrases from target hypotheses and form a syn- chronous context-free grammar to guide a series of transformations of target hy- potheses into fused translations. The model is able to exploit phrasal and struc- tural system-weighted consensus and also to utilize existing information about word ordering present in the target hypotheses. In addition, to consider a diverse set of plausible fused translations, we develop a hybrid combination architecture, where we paraphrase every target hypothesis us- ing different fusing techniques to obtain fused translations for each target, and then make the final selection among all fused translations. Our experimental re- sults show that our approach can achieve a significant improvement over combina- tion baselines. 1 Introduction In the past several years, many machine transla- tion (MT) combination approaches have been developed. Word-level combination approaches, such as the confusion network decoding model, have been quite successful (Matusov et al., 2006; Rosti et al., 2007a; He et al. 2008; Karakos et al. 2008; Chen et al. 2009a; Narsale 2010; Leusch 2011; Freitag et al. 2014). In addition to word-level combination ap- proaches, some phrase-level combination ap- proaches have also recently been developed; the goal is to retain coherence and consistency be- tween the words in a phrase. The most common phrase-level combination approaches are re- decoding methods: by constructing a new phrase table from each MT system’s source-to-target phrase alignments, the source sentence can also be re-decoded using the new translation table (Rosti et al., 2007b; Huang and Papineni, 2007; Chen et al., 2007; Chen et al., 2009b). One prob- lem with these approaches is that, just with a new phrase table, existing information about word ordering present in the target hypotheses is not utilized; thus the approaches are likely to make new mistakes of word reordering which do not appear in the target hypotheses of MT engines. Huang and Papineni (2007) attacked this issue through a reordering cost function that encour- ages search along with decoding paths from all MT engines’ decoders. Another phrase-level combination approach relies on a lattice decoding model to carry out the combination (Feng et al 2009; Du and Way 2010; Ma and McKeown 2012). In a lattice, each edge is associated with a phrase (a single word or a sequence of words) rather than a single word. The construction of the lattice is based on the extraction of phrase pairs from word alignments between a selected best MT system hypothesis (the backbone) and the other translation hypothe- ses. One challenge of the lattice decoding model is …

Content truncated to 500 words