Learning to translate in real-time with neural machine translation

Jiatao Gu, Graham Neubig, Kyunghyun Cho, Victor O.K. Li

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

Abstract

Translating in real-time, a.k.a. simultaneous translation, outputs translation words before the input sentence ends, which is a challenging problem for conventional machine translation methods. We propose a neural machine translation (NMT) framework for simultaneous translation in which an agent learns to make decisions on when to translate from the interaction with a pre-trained NMT environment. To trade off quality and delay, we extensively explore various targets for delay and design a method for beam-search applicable in the simultaneous MT setting. Experiments against state-of-the-art baselines on two language pairs demonstrate the efficacy of the proposed framework both quantitatively and qualitatively.

Original languageEnglish (US)
Title of host publicationLong Papers - Continued
PublisherAssociation for Computational Linguistics (ACL)
Pages1053-1062
Number of pages10
Volume1
ISBN (Electronic)9781510838604
StatePublished - 2017
Event15th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2017 - Valencia, Spain
Duration: Apr 3 2017Apr 7 2017

Other

Other15th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2017
CountrySpain
CityValencia
Period4/3/174/7/17

Fingerprint

learning
time
Machine Translation
experiment
interaction
language
Efficacy
Translating
Experiment
Interaction
Translation Method
Conventional
Language

ASJC Scopus subject areas

  • Linguistics and Language
  • Language and Linguistics

Cite this

Gu, J., Neubig, G., Cho, K., & Li, V. O. K. (2017). Learning to translate in real-time with neural machine translation. In Long Papers - Continued (Vol. 1, pp. 1053-1062). Association for Computational Linguistics (ACL).

Learning to translate in real-time with neural machine translation. / Gu, Jiatao; Neubig, Graham; Cho, Kyunghyun; Li, Victor O.K.

Long Papers - Continued. Vol. 1 Association for Computational Linguistics (ACL), 2017. p. 1053-1062.

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

Gu, J, Neubig, G, Cho, K & Li, VOK 2017, Learning to translate in real-time with neural machine translation. in Long Papers - Continued. vol. 1, Association for Computational Linguistics (ACL), pp. 1053-1062, 15th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2017, Valencia, Spain, 4/3/17.
Gu J, Neubig G, Cho K, Li VOK. Learning to translate in real-time with neural machine translation. In Long Papers - Continued. Vol. 1. Association for Computational Linguistics (ACL). 2017. p. 1053-1062
Gu, Jiatao ; Neubig, Graham ; Cho, Kyunghyun ; Li, Victor O.K. / Learning to translate in real-time with neural machine translation. Long Papers - Continued. Vol. 1 Association for Computational Linguistics (ACL), 2017. pp. 1053-1062
@inproceedings{1753016ae3f8429e9a8aa43f73fbc92c,
title = "Learning to translate in real-time with neural machine translation",
abstract = "Translating in real-time, a.k.a. simultaneous translation, outputs translation words before the input sentence ends, which is a challenging problem for conventional machine translation methods. We propose a neural machine translation (NMT) framework for simultaneous translation in which an agent learns to make decisions on when to translate from the interaction with a pre-trained NMT environment. To trade off quality and delay, we extensively explore various targets for delay and design a method for beam-search applicable in the simultaneous MT setting. Experiments against state-of-the-art baselines on two language pairs demonstrate the efficacy of the proposed framework both quantitatively and qualitatively.",
author = "Jiatao Gu and Graham Neubig and Kyunghyun Cho and Li, {Victor O.K.}",
year = "2017",
language = "English (US)",
volume = "1",
pages = "1053--1062",
booktitle = "Long Papers - Continued",
publisher = "Association for Computational Linguistics (ACL)",

}

TY - GEN

T1 - Learning to translate in real-time with neural machine translation

AU - Gu, Jiatao

AU - Neubig, Graham

AU - Cho, Kyunghyun

AU - Li, Victor O.K.

PY - 2017

Y1 - 2017

N2 - Translating in real-time, a.k.a. simultaneous translation, outputs translation words before the input sentence ends, which is a challenging problem for conventional machine translation methods. We propose a neural machine translation (NMT) framework for simultaneous translation in which an agent learns to make decisions on when to translate from the interaction with a pre-trained NMT environment. To trade off quality and delay, we extensively explore various targets for delay and design a method for beam-search applicable in the simultaneous MT setting. Experiments against state-of-the-art baselines on two language pairs demonstrate the efficacy of the proposed framework both quantitatively and qualitatively.

AB - Translating in real-time, a.k.a. simultaneous translation, outputs translation words before the input sentence ends, which is a challenging problem for conventional machine translation methods. We propose a neural machine translation (NMT) framework for simultaneous translation in which an agent learns to make decisions on when to translate from the interaction with a pre-trained NMT environment. To trade off quality and delay, we extensively explore various targets for delay and design a method for beam-search applicable in the simultaneous MT setting. Experiments against state-of-the-art baselines on two language pairs demonstrate the efficacy of the proposed framework both quantitatively and qualitatively.

UR - http://www.scopus.com/inward/record.url?scp=85021624692&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021624692&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1

SP - 1053

EP - 1062

BT - Long Papers - Continued

PB - Association for Computational Linguistics (ACL)

ER -