Skip to main navigation Skip to search Skip to main content

Neural machine translation with phrasal attention

  • Yachao Li*
  • , Deyi Xiong
  • , Min Zhang
  • *Corresponding author for this work
  • Soochow University
  • Northwest Minzu University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Attention-based neural machine translation (NMT) employs an attention network to capture structural correspondences between the source and target language at the word level. Unfortunately, alignments between source and target equivalents are complicated, which makes word-level attention not adequate to model these relations (e.g., alignments between a source idiom and its target translation). In order to handle this issue, we propose a phrase-level attention mechanism to complement the word-level attention network in this paper. The proposed phrasal attention framework is simple yet effective, keeping the strength of phrase-based statistical machine translation (SMT) on the source side. Experiments on Chinese-to-English translation task demonstrate that the proposed method is able to statistically improve word-level attention-based NMT.

Original languageEnglish
Title of host publicationMachine Translation - 13th China Workshop, CWMT 2017, Revised Selected Papers
EditorsDerek F. Wong, Deyi Xiong
PublisherSpringer Verlag
Pages1-8
Number of pages8
ISBN (Print)9789811071331
DOIs
StatePublished - 2017
Externally publishedYes
Event13th China Workshop on Machine Translation, CWMT 2017 - Dalian, China
Duration: 27 Sep 201729 Sep 2017

Publication series

NameCommunications in Computer and Information Science
Volume787
ISSN (Print)1865-0929

Conference

Conference13th China Workshop on Machine Translation, CWMT 2017
Country/TerritoryChina
CityDalian
Period27/09/1729/09/17

Keywords

  • Attention mechaism
  • Gated recurrent unit
  • Neural machine translation
  • Recurrent neural network

Fingerprint

Dive into the research topics of 'Neural machine translation with phrasal attention'. Together they form a unique fingerprint.

Cite this