On integrating a language model into neural machine translation

Gulcehre C., Firat O., Xu K., Cho K., Bengio Y.

COMPUTER SPEECH AND LANGUAGE, vol.45, pp.137-148, 2017 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 45
  • Publication Date: 2017
  • Doi Number: 10.1016/j.csl.2017.01.014
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.137-148
  • Keywords: Neural machine translation, Monolingual data, Language models, Low resource machine translation, Deep learning, Neural network
  • Middle East Technical University Affiliated: Yes


Recent advances in end-to-end neural machine translation models have achieved promising results on high-resource language pairs such as En -> Fr and En -> De. One of the major factor behind these successes is the availability of high quality parallel corpora. We explore two strategies on leveraging abundant amount of monolingual data for neural machine translation. We observe improvements by both combining scores from neural language model trained only on target monolingual data with neural machine translation model and fusing hidden-states of these two models. We obtain up to 2 BLEU improvement over hierarchical and phrase-based baseline on low-resource language pair, Turkish -> English. Our method was initially motivated towards tasks with less parallel data, but we also show that it extends to high resource languages such as Cs -> En and De -> En translation tasks, where we obtain 0.39 and 0.47 BLEU improvements over the neural machine translation baselines, respectively. (C) 2017 Elsevier Ltd. All rights reserved.