Neural Lattice Search for Domain Adaptation in Machine Translation

Huda Khayrallah, Gaurav Kumar, Kevin Duh, Matt Post, Philipp Koehn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Domain adaptation is a major challenge for neural machine translation (NMT). Given unknown words or new domains, NMT systems tend to generate fluent translations at the expense of adequacy. We present a stack-based lattice search algorithm for NMT and show that constraining its search space with lattices generated by phrase-based machine translation (PBMT) improves robustness. We report consistent BLEU score gains across four diverse domain adaptation tasks involving medical, IT, Koran, or subtitles texts.
Original languageEnglish
Title of host publicationProceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 2: Short Papers)
Place of PublicationTaipei, Taiwan
PublisherAsian Federation of Natural Language Processing
Pages20-25
Number of pages6
Publication statusPublished - 1 Dec 2017
EventThe 8th International Joint Conference on Natural Language Processing - Taipei, Taiwan, Province of China
Duration: 27 Nov 20171 Dec 2017
http://ijcnlp2017.org/site/page.aspx?pid=901&sid=1133&lang=en

Conference

ConferenceThe 8th International Joint Conference on Natural Language Processing
Abbreviated titleIJCNLP 2017
CountryTaiwan, Province of China
CityTaipei
Period27/11/171/12/17
Internet address

Cite this