A deep fusion model for domain adaptation in phrase-based MT

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

1 Citation (Scopus)

Abstract

We present a novel fusion model for domain adaptation in Statistical Machine Translation. Our model is based on the joint source-target neural network (Devlin et al., 2014), and is learned by fusing in- and out-domain models. The adaptation is performed by backpropagating errors from the output layer to the word embedding layer of each model, subsequently adjusting parameters of the composite model towards the in-domain data. On the standard tasks of translating English-to-German and Arabic-to-English TED talks, we observed average improvements of +0.9 and +0.7 BLEU points, respectively over a competition grade phrase-based system. We also demonstrate improvements over existing adaptation methods.

Original languageEnglish
Title of host publicationCOLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016
Subtitle of host publicationTechnical Papers
PublisherAssociation for Computational Linguistics, ACL Anthology
Pages3177-3187
Number of pages11
ISBN (Print)9784879747020
Publication statusPublished - 1 Jan 2016
Event26th International Conference on Computational Linguistics, COLING 2016 - Osaka, Japan
Duration: 11 Dec 201616 Dec 2016

Other

Other26th International Conference on Computational Linguistics, COLING 2016
CountryJapan
CityOsaka
Period11/12/1616/12/16

Fingerprint

Fusion reactions
neural network
Fusion
Neural networks
Composite materials
Layer

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Language and Linguistics
  • Linguistics and Language

Cite this

Durrani, N., Sajjad, H., Rayhan Joty, S., & Abdelali, A. (2016). A deep fusion model for domain adaptation in phrase-based MT. In COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers (pp. 3177-3187). Association for Computational Linguistics, ACL Anthology.

A deep fusion model for domain adaptation in phrase-based MT. / Durrani, Nadir; Sajjad, Hassan; Rayhan Joty, Shafiq; Abdelali, Ahmed.

COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers. Association for Computational Linguistics, ACL Anthology, 2016. p. 3177-3187.

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

Durrani, N, Sajjad, H, Rayhan Joty, S & Abdelali, A 2016, A deep fusion model for domain adaptation in phrase-based MT. in COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers. Association for Computational Linguistics, ACL Anthology, pp. 3177-3187, 26th International Conference on Computational Linguistics, COLING 2016, Osaka, Japan, 11/12/16.
Durrani N, Sajjad H, Rayhan Joty S, Abdelali A. A deep fusion model for domain adaptation in phrase-based MT. In COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers. Association for Computational Linguistics, ACL Anthology. 2016. p. 3177-3187
Durrani, Nadir ; Sajjad, Hassan ; Rayhan Joty, Shafiq ; Abdelali, Ahmed. / A deep fusion model for domain adaptation in phrase-based MT. COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016: Technical Papers. Association for Computational Linguistics, ACL Anthology, 2016. pp. 3177-3187
@inproceedings{8160fcc0bd5a454991aa74c2ae62b703,
title = "A deep fusion model for domain adaptation in phrase-based MT",
abstract = "We present a novel fusion model for domain adaptation in Statistical Machine Translation. Our model is based on the joint source-target neural network (Devlin et al., 2014), and is learned by fusing in- and out-domain models. The adaptation is performed by backpropagating errors from the output layer to the word embedding layer of each model, subsequently adjusting parameters of the composite model towards the in-domain data. On the standard tasks of translating English-to-German and Arabic-to-English TED talks, we observed average improvements of +0.9 and +0.7 BLEU points, respectively over a competition grade phrase-based system. We also demonstrate improvements over existing adaptation methods.",
author = "Nadir Durrani and Hassan Sajjad and {Rayhan Joty}, Shafiq and Ahmed Abdelali",
year = "2016",
month = "1",
day = "1",
language = "English",
isbn = "9784879747020",
pages = "3177--3187",
booktitle = "COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016",
publisher = "Association for Computational Linguistics, ACL Anthology",

}

TY - GEN

T1 - A deep fusion model for domain adaptation in phrase-based MT

AU - Durrani, Nadir

AU - Sajjad, Hassan

AU - Rayhan Joty, Shafiq

AU - Abdelali, Ahmed

PY - 2016/1/1

Y1 - 2016/1/1

N2 - We present a novel fusion model for domain adaptation in Statistical Machine Translation. Our model is based on the joint source-target neural network (Devlin et al., 2014), and is learned by fusing in- and out-domain models. The adaptation is performed by backpropagating errors from the output layer to the word embedding layer of each model, subsequently adjusting parameters of the composite model towards the in-domain data. On the standard tasks of translating English-to-German and Arabic-to-English TED talks, we observed average improvements of +0.9 and +0.7 BLEU points, respectively over a competition grade phrase-based system. We also demonstrate improvements over existing adaptation methods.

AB - We present a novel fusion model for domain adaptation in Statistical Machine Translation. Our model is based on the joint source-target neural network (Devlin et al., 2014), and is learned by fusing in- and out-domain models. The adaptation is performed by backpropagating errors from the output layer to the word embedding layer of each model, subsequently adjusting parameters of the composite model towards the in-domain data. On the standard tasks of translating English-to-German and Arabic-to-English TED talks, we observed average improvements of +0.9 and +0.7 BLEU points, respectively over a competition grade phrase-based system. We also demonstrate improvements over existing adaptation methods.

UR - http://www.scopus.com/inward/record.url?scp=85049126112&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049126112&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9784879747020

SP - 3177

EP - 3187

BT - COLING 2016 - 26th International Conference on Computational Linguistics, Proceedings of COLING 2016

PB - Association for Computational Linguistics, ACL Anthology

ER -