Skip to main navigation Skip to search Skip to main content

Pivot-based semantic splicing for neural machine translation

  • Harbin Institute of Technology

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

Abstract

Current neural machine translation (NMT) usually extracts a fixedlength semantic representation for source sentence, and then depends on this representation to generate corresponding target translation. In this paper, we proposed a pivot-based semantic splicing model (PBSSM) to obtain a semantic representation including more translation information for source sentence, thus improving the translation performance of NMT. The spliced semantic representation is derived from source languages of trilingual parallel corpus by the pivot-based NMT. Besides, the proposed PBSSM only depends on one source language to generate its semantic representation during the encoding process. We integrated it into the NMT architecture. Experiments on the English-Japanese translation task show that our model achieves a substantial improvement by up to 22.9% (3.74 BLEU) over the baseline.

Original languageEnglish
Title of host publicationMachine Translation - 12th China Workshop, CWMT 2016, Revised Selected Papers
EditorsShujie Liu, Muyun Yang
PublisherSpringer Verlag
Pages14-24
Number of pages11
ISBN (Print)9789811036347
DOIs
StatePublished - 2016
Event12th China Workshop on Machine Translation, CWMT 2016 - Urumqi, China
Duration: 25 Aug 201626 Aug 2016

Publication series

NameCommunications in Computer and Information Science
Volume668
ISSN (Print)1865-0929

Conference

Conference12th China Workshop on Machine Translation, CWMT 2016
Country/TerritoryChina
CityUrumqi
Period25/08/1626/08/16

Keywords

  • Neural machine translation
  • Pivot-based translation
  • Semantic splicing

Fingerprint

Dive into the research topics of 'Pivot-based semantic splicing for neural machine translation'. Together they form a unique fingerprint.

Cite this