InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation

Gabriele Sarti; Arianna Bisazza

InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation

Gabriele Sarti, Arianna Bisazza

Computational Linguistics (CL)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

40 Downloads (Pure)

Abstract

Neural machine translation (NMT) systems are nowadays essential components of professional translation workflows. Consequently, human translators are increasingly working as post-editors for machine-translated content. The NWO-funded InDeep project aims to empower users of Deep Learning models of text, speech, and music by improving their ability to interact with such models and interpret their behaviors. In the specific context of translation, we aim at developing new tools and methodologies to improve prediction attribution, error analysis, and controllable generation for NMT systems. These advances will be evaluated through field studies involving professional translators to assess gains in efficiency and overall enjoyability of the post-editing process.

Original language	English
Title of host publication	Proceedings of the 23rd Annual Conference of the European Association for Machine Translation
Editors	Helena Moniz, Lieve Macken, Andrew Rufener, Loïc Barrault, Marta R. Costa-jussà, Christophe Declercq, Maarit Koponen, Ellie Kemp, Spyridon Pilos, Mikel L. Forcada, Carolina Scarton, Joachim van den Bogaert, Joke Daems, Arda Tezcan, Bram Vanroy, Margot Fonteyne
Publisher	European Association for Machine Translation
Pages	313-314
Number of pages	2
Publication status	Published - Jun-2022
Event	23rd Annual Conference of the European Association for Machine Translation - Zebrastraat, Ghent, Belgium Duration: 1-Jun-2022 → 3-Jun-2022 https://eamt2022.com/

Conference

Conference	23rd Annual Conference of the European Association for Machine Translation
Abbreviated title	EAMT '22
Country/Territory	Belgium
City	Ghent
Period	01/06/2022 → 03/06/2022
Internet address	https://eamt2022.com/

Keywords

machine translation
interpretability

Access to Document

InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine TranslationFinal publisher's version, 72.2 KBLicence: CC BY-ND

Handle.net

Persistent link

1 Preprint

Inseq: An Interpretability Toolkit for Sequence Generation Models
Sarti, G., Feldhus, N., Sickert, L., Wal, O. V. D., Nissim, M. & Bisazza, A., 27-Feb-2023, (Submitted) arXiv, 15 p.
Research output: Working paper › Preprint › Academic

File
102 Downloads (Pure)

1 Hosting a visitor
1 Professional or public presentation

Towards User-centric Interpretability of NLP Models
Sarti, G. (Speaker)
18-May-2022
Activity: Talk and presentation › Professional or public presentation › Professional
Michael Carl
Sarti, G. (Host) & Bisazza, A. (Host)
23-May-2022 → 30-May-2022
Activity: Hosting a visitor › Academic

Cite this

Sarti, G., & Bisazza, A. (2022). InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation. In H. Moniz, L. Macken, A. Rufener, L. Barrault, M. R. Costa-jussà, C. Declercq, M. Koponen, E. Kemp, S. Pilos, M. L. Forcada, C. Scarton, J. van den Bogaert, J. Daems, A. Tezcan, B. Vanroy, & M. Fonteyne (Eds.), Proceedings of the 23rd Annual Conference of the European Association for Machine Translation (pp. 313-314). European Association for Machine Translation.

Sarti, Gabriele ; Bisazza, Arianna. / InDeep × NMT : Empowering Human Translators via Interpretable Neural Machine Translation. Proceedings of the 23rd Annual Conference of the European Association for Machine Translation. editor / Helena Moniz ; Lieve Macken ; Andrew Rufener ; Loïc Barrault ; Marta R. Costa-jussà ; Christophe Declercq ; Maarit Koponen ; Ellie Kemp ; Spyridon Pilos ; Mikel L. Forcada ; Carolina Scarton ; Joachim van den Bogaert ; Joke Daems ; Arda Tezcan ; Bram Vanroy ; Margot Fonteyne. European Association for Machine Translation, 2022. pp. 313-314

@inproceedings{7e8a119594314560851717ee3b99e808,

title = "InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation",

abstract = "Neural machine translation (NMT) systems are nowadays essential components of professional translation workflows. Consequently, human translators are increasingly working as post-editors for machine-translated content. The NWO-funded InDeep project aims to empower users of Deep Learning models of text, speech, and music by improving their ability to interact with such models and interpret their behaviors. In the specific context of translation, we aim at developing new tools and methodologies to improve prediction attribution, error analysis, and controllable generation for NMT systems. These advances will be evaluated through field studies involving professional translators to assess gains in efficiency and overall enjoyability of the post-editing process.",

keywords = "machine translation, interpretability",

author = "Gabriele Sarti and Arianna Bisazza",

year = "2022",

month = jun,

language = "English",

pages = "313--314",

editor = "Helena Moniz and Lieve Macken and Andrew Rufener and Lo{\"i}c Barrault and Costa-juss{\`a}, {Marta R.} and Christophe Declercq and Maarit Koponen and Ellie Kemp and Spyridon Pilos and Forcada, {Mikel L.} and Carolina Scarton and {van den Bogaert}, Joachim and Joke Daems and Arda Tezcan and Bram Vanroy and Margot Fonteyne",

booktitle = "Proceedings of the 23rd Annual Conference of the European Association for Machine Translation",

publisher = "European Association for Machine Translation",

note = "23rd Annual Conference of the European Association for Machine Translation, EAMT '22 ; Conference date: 01-06-2022 Through 03-06-2022",

url = "https://eamt2022.com/",

}

Sarti, G & Bisazza, A 2022, InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation. in H Moniz, L Macken, A Rufener, L Barrault, MR Costa-jussà, C Declercq, M Koponen, E Kemp, S Pilos, ML Forcada, C Scarton, J van den Bogaert, J Daems, A Tezcan, B Vanroy & M Fonteyne (eds), Proceedings of the 23rd Annual Conference of the European Association for Machine Translation. European Association for Machine Translation, pp. 313-314, 23rd Annual Conference of the European Association for Machine Translation, Ghent, Belgium, 01/06/2022.

InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation. / Sarti, Gabriele ; Bisazza, Arianna.
Proceedings of the 23rd Annual Conference of the European Association for Machine Translation. ed. / Helena Moniz; Lieve Macken; Andrew Rufener; Loïc Barrault; Marta R. Costa-jussà; Christophe Declercq; Maarit Koponen; Ellie Kemp; Spyridon Pilos; Mikel L. Forcada; Carolina Scarton; Joachim van den Bogaert; Joke Daems; Arda Tezcan; Bram Vanroy; Margot Fonteyne. European Association for Machine Translation, 2022. p. 313-314.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - InDeep × NMT

T2 - 23rd Annual Conference of the European Association for Machine Translation

AU - Sarti, Gabriele

AU - Bisazza, Arianna

PY - 2022/6

Y1 - 2022/6

N2 - Neural machine translation (NMT) systems are nowadays essential components of professional translation workflows. Consequently, human translators are increasingly working as post-editors for machine-translated content. The NWO-funded InDeep project aims to empower users of Deep Learning models of text, speech, and music by improving their ability to interact with such models and interpret their behaviors. In the specific context of translation, we aim at developing new tools and methodologies to improve prediction attribution, error analysis, and controllable generation for NMT systems. These advances will be evaluated through field studies involving professional translators to assess gains in efficiency and overall enjoyability of the post-editing process.

AB - Neural machine translation (NMT) systems are nowadays essential components of professional translation workflows. Consequently, human translators are increasingly working as post-editors for machine-translated content. The NWO-funded InDeep project aims to empower users of Deep Learning models of text, speech, and music by improving their ability to interact with such models and interpret their behaviors. In the specific context of translation, we aim at developing new tools and methodologies to improve prediction attribution, error analysis, and controllable generation for NMT systems. These advances will be evaluated through field studies involving professional translators to assess gains in efficiency and overall enjoyability of the post-editing process.

KW - machine translation

KW - interpretability

M3 - Conference contribution

SP - 313

EP - 314

BT - Proceedings of the 23rd Annual Conference of the European Association for Machine Translation

A2 - Moniz, Helena

A2 - Macken, Lieve

A2 - Rufener, Andrew

A2 - Barrault, Loïc

A2 - Costa-jussà, Marta R.

A2 - Declercq, Christophe

A2 - Koponen, Maarit

A2 - Kemp, Ellie

A2 - Pilos, Spyridon

A2 - Forcada, Mikel L.

A2 - Scarton, Carolina

A2 - van den Bogaert, Joachim

A2 - Daems, Joke

A2 - Tezcan, Arda

A2 - Vanroy, Bram

A2 - Fonteyne, Margot

PB - European Association for Machine Translation

Y2 - 1 June 2022 through 3 June 2022

ER -

Sarti G , Bisazza A. InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation. In Moniz H, Macken L, Rufener A, Barrault L, Costa-jussà MR, Declercq C, Koponen M, Kemp E, Pilos S, Forcada ML, Scarton C, van den Bogaert J, Daems J, Tezcan A, Vanroy B, Fonteyne M, editors, Proceedings of the 23rd Annual Conference of the European Association for Machine Translation. European Association for Machine Translation. 2022. p. 313-314

InDeep × NMT: Empowering Human Translators via Interpretable Neural Machine Translation

Abstract

Conference

Keywords

Access to Document

Handle.net

Fingerprint

Research output

Inseq: An Interpretability Toolkit for Sequence Generation Models

Activities

Towards User-centric Interpretability of NLP Models

Michael Carl

Cite this