The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT)

Faheem Kirefu; Vivek Iyer; Pinzhen Chen; Laurie Burchell

The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT)

Faheem Kirefu, Vivek Iyer, Pinzhen Chen, Laurie Burchell

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

The University of Edinburgh participated in the WMT22 shared task on code-mixed translation. This consists of two subtasks: i) generating code-mixed Hindi/English (Hinglish) text generation from parallel Hindi and English sentences and ii) machine translation from Hinglish to English. As both subtasks are considered low-resource, we focused our efforts on careful data generation and curation, especially the use of backtranslation from monolingual resources. For subtask 1 we explored the effects of constrained decoding on English and transliterated subwords in order to produce Hinglish. For subtask 2, we investigated different pretraining techniques, namely comparing simple initialisation from existing machine translation models and aligned augmentation. For both subtasks, we found that our baseline systems worked best. Our systems for both subtasks were one of the overall top-performing submissions.

Original language	English
Title of host publication	Proceedings of the Seventh Conference on Machine Translation
Editors	Philipp Koehn, Loïc Barrault, Ondřej Bojar, Fethi Bougares, Rajen Chatterjee, Marta R. Costa-jussà, Christian Federmann, Mark Fishel, Alexander Fraser, Markus Freitag, Yvette Graham, Roman Grundkiewicz, Paco Guzman, Barry Haddow, Matthias Huck, Antonio Jimeno Yepes, Tom Kocmi, André Martins, Makoto Morishita, Christof Monz, Masaaki Nagata, Toshiaki Nakazawa, Matteo Negri, Aurélie Névéol, Mariana Neves, Martin Popel, Marco Turchi, Marcos Zampieri
Place of Publication	Abu Dhabi, United Arab Emirates
Publisher	Association for Computational Linguistics
Pages	1145-1157
Number of pages	13
ISBN (Electronic)	9781959429296
Publication status	Published - 1 Dec 2022
Event	Seventh Conference on Machine Translation - Abu Dhabi, United Arab Emirates Duration: 7 Dec 2022 → 8 Dec 2022 Conference number: 7 https://statmt.org/wmt22/

Publication series

Name	Proceedings of the Conference on Machine Translation
Publisher	ACL
ISSN (Electronic)	2768-0983

Conference

Conference	Seventh Conference on Machine Translation
Abbreviated title	WMT22
Country/Territory	United Arab Emirates
City	Abu Dhabi
Period	7/12/22 → 8/12/22
Internet address	https://statmt.org/wmt22/

Access to Document

The University of Edinburgh's_KIREFU_DOA09102022_VOR_CC BYFinal published version, 208 KBLicence: Creative Commons: Attribution (CC-BY)

https://aclanthology.org/2022.wmt-1.115Licence: Creative Commons: Attribution (CC-BY)

Cite this

Kirefu, F., Iyer, V., Chen, P., & Burchell, L. (2022). The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT). In P. Koehn, L. Barrault, O. Bojar, F. Bougares, R. Chatterjee, M. R. Costa-jussà, C. Federmann, M. Fishel, A. Fraser, M. Freitag, Y. Graham, R. Grundkiewicz, P. Guzman, B. Haddow, M. Huck, A. Jimeno Yepes, T. Kocmi, A. Martins, M. Morishita, C. Monz, M. Nagata, T. Nakazawa, M. Negri, A. Névéol, M. Neves, M. Popel, M. Turchi, ... M. Zampieri (Eds.), Proceedings of the Seventh Conference on Machine Translation (pp. 1145-1157). (Proceedings of the Conference on Machine Translation). Association for Computational Linguistics. https://aclanthology.org/2022.wmt-1.115

Kirefu, Faheem ; Iyer, Vivek ; Chen, Pinzhen et al. / The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT). Proceedings of the Seventh Conference on Machine Translation. editor / Philipp Koehn ; Loïc Barrault ; Ondřej Bojar ; Fethi Bougares ; Rajen Chatterjee ; Marta R. Costa-jussà ; Christian Federmann ; Mark Fishel ; Alexander Fraser ; Markus Freitag ; Yvette Graham ; Roman Grundkiewicz ; Paco Guzman ; Barry Haddow ; Matthias Huck ; Antonio Jimeno Yepes ; Tom Kocmi ; André Martins ; Makoto Morishita ; Christof Monz ; Masaaki Nagata ; Toshiaki Nakazawa ; Matteo Negri ; Aurélie Névéol ; Mariana Neves ; Martin Popel ; Marco Turchi ; Marcos Zampieri. Abu Dhabi, United Arab Emirates : Association for Computational Linguistics, 2022. pp. 1145-1157 (Proceedings of the Conference on Machine Translation).

@inproceedings{051b427239bd42f4890bc2261010a2e0,

title = "The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT)",

abstract = "The University of Edinburgh participated in the WMT22 shared task on code-mixed translation. This consists of two subtasks: i) generating code-mixed Hindi/English (Hinglish) text generation from parallel Hindi and English sentences and ii) machine translation from Hinglish to English. As both subtasks are considered low-resource, we focused our efforts on careful data generation and curation, especially the use of backtranslation from monolingual resources. For subtask 1 we explored the effects of constrained decoding on English and transliterated subwords in order to produce Hinglish. For subtask 2, we investigated different pretraining techniques, namely comparing simple initialisation from existing machine translation models and aligned augmentation. For both subtasks, we found that our baseline systems worked best. Our systems for both subtasks were one of the overall top-performing submissions.",

author = "Faheem Kirefu and Vivek Iyer and Pinzhen Chen and Laurie Burchell",

year = "2022",

month = dec,

day = "1",

language = "English",

series = "Proceedings of the Conference on Machine Translation",

publisher = "Association for Computational Linguistics",

pages = "1145--1157",

editor = "Philipp Koehn and Barrault, { Lo{\"i}c} and Ond{\v r}ej Bojar and Fethi Bougares and Rajen Chatterjee and Costa-juss{\`a}, {Marta R.} and Christian Federmann and Mark Fishel and Alexander Fraser and Markus Freitag and Yvette Graham and Roman Grundkiewicz and Paco Guzman and Barry Haddow and Matthias Huck and {Jimeno Yepes}, Antonio and Tom Kocmi and Andr{\'e} Martins and Makoto Morishita and Christof Monz and Masaaki Nagata and Toshiaki Nakazawa and Matteo Negri and Aur{\'e}lie N{\'e}v{\'e}ol and Mariana Neves and Martin Popel and Marco Turchi and Marcos Zampieri",

booktitle = "Proceedings of the Seventh Conference on Machine Translation",

note = "Seventh Conference on Machine Translation, WMT22 ; Conference date: 07-12-2022 Through 08-12-2022",

url = "https://statmt.org/wmt22/",

}

Kirefu, F, Iyer, V, Chen, P & Burchell, L 2022, The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT). in P Koehn, L Barrault, O Bojar, F Bougares, R Chatterjee, MR Costa-jussà, C Federmann, M Fishel, A Fraser, M Freitag, Y Graham, R Grundkiewicz, P Guzman, B Haddow, M Huck, A Jimeno Yepes, T Kocmi, A Martins, M Morishita, C Monz, M Nagata, T Nakazawa, M Negri, A Névéol, M Neves, M Popel, M Turchi & M Zampieri (eds), Proceedings of the Seventh Conference on Machine Translation. Proceedings of the Conference on Machine Translation, Association for Computational Linguistics, Abu Dhabi, United Arab Emirates, pp. 1145-1157, Seventh Conference on Machine Translation, Abu Dhabi, United Arab Emirates, 7/12/22. <https://aclanthology.org/2022.wmt-1.115>

The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT). / Kirefu, Faheem; Iyer, Vivek; Chen, Pinzhen et al.
Proceedings of the Seventh Conference on Machine Translation. ed. / Philipp Koehn; Loïc Barrault; Ondřej Bojar; Fethi Bougares; Rajen Chatterjee; Marta R. Costa-jussà; Christian Federmann; Mark Fishel; Alexander Fraser; Markus Freitag; Yvette Graham; Roman Grundkiewicz; Paco Guzman; Barry Haddow; Matthias Huck; Antonio Jimeno Yepes; Tom Kocmi; André Martins; Makoto Morishita; Christof Monz; Masaaki Nagata; Toshiaki Nakazawa; Matteo Negri; Aurélie Névéol; Mariana Neves; Martin Popel; Marco Turchi; Marcos Zampieri. Abu Dhabi, United Arab Emirates: Association for Computational Linguistics, 2022. p. 1145-1157 (Proceedings of the Conference on Machine Translation).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT)

AU - Kirefu, Faheem

AU - Iyer, Vivek

AU - Chen, Pinzhen

AU - Burchell, Laurie

N1 - Conference code: 7

PY - 2022/12/1

Y1 - 2022/12/1

N2 - The University of Edinburgh participated in the WMT22 shared task on code-mixed translation. This consists of two subtasks: i) generating code-mixed Hindi/English (Hinglish) text generation from parallel Hindi and English sentences and ii) machine translation from Hinglish to English. As both subtasks are considered low-resource, we focused our efforts on careful data generation and curation, especially the use of backtranslation from monolingual resources. For subtask 1 we explored the effects of constrained decoding on English and transliterated subwords in order to produce Hinglish. For subtask 2, we investigated different pretraining techniques, namely comparing simple initialisation from existing machine translation models and aligned augmentation. For both subtasks, we found that our baseline systems worked best. Our systems for both subtasks were one of the overall top-performing submissions.

AB - The University of Edinburgh participated in the WMT22 shared task on code-mixed translation. This consists of two subtasks: i) generating code-mixed Hindi/English (Hinglish) text generation from parallel Hindi and English sentences and ii) machine translation from Hinglish to English. As both subtasks are considered low-resource, we focused our efforts on careful data generation and curation, especially the use of backtranslation from monolingual resources. For subtask 1 we explored the effects of constrained decoding on English and transliterated subwords in order to produce Hinglish. For subtask 2, we investigated different pretraining techniques, namely comparing simple initialisation from existing machine translation models and aligned augmentation. For both subtasks, we found that our baseline systems worked best. Our systems for both subtasks were one of the overall top-performing submissions.

M3 - Conference contribution

T3 - Proceedings of the Conference on Machine Translation

SP - 1145

EP - 1157

BT - Proceedings of the Seventh Conference on Machine Translation

A2 - Koehn, Philipp

A2 - Barrault, Loïc

A2 - Bojar, Ondřej

A2 - Bougares, Fethi

A2 - Chatterjee, Rajen

A2 - Costa-jussà, Marta R.

A2 - Federmann, Christian

A2 - Fishel, Mark

A2 - Fraser, Alexander

A2 - Freitag, Markus

A2 - Graham, Yvette

A2 - Grundkiewicz, Roman

A2 - Guzman, Paco

A2 - Haddow, Barry

A2 - Huck, Matthias

A2 - Jimeno Yepes, Antonio

A2 - Kocmi, Tom

A2 - Martins, André

A2 - Morishita, Makoto

A2 - Monz, Christof

A2 - Nagata, Masaaki

A2 - Nakazawa, Toshiaki

A2 - Negri, Matteo

A2 - Névéol, Aurélie

A2 - Neves, Mariana

A2 - Popel, Martin

A2 - Turchi, Marco

A2 - Zampieri, Marcos

PB - Association for Computational Linguistics

CY - Abu Dhabi, United Arab Emirates

T2 - Seventh Conference on Machine Translation

Y2 - 7 December 2022 through 8 December 2022

ER -

Kirefu F, Iyer V, Chen P , Burchell L. The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT). In Koehn P, Barrault L, Bojar O, Bougares F, Chatterjee R, Costa-jussà MR, Federmann C, Fishel M, Fraser A, Freitag M, Graham Y, Grundkiewicz R, Guzman P, Haddow B, Huck M, Jimeno Yepes A, Kocmi T, Martins A, Morishita M, Monz C, Nagata M, Nakazawa T, Negri M, Névéol A, Neves M, Popel M, Turchi M, Zampieri M, editors, Proceedings of the Seventh Conference on Machine Translation. Abu Dhabi, United Arab Emirates: Association for Computational Linguistics. 2022. p. 1145-1157. (Proceedings of the Conference on Machine Translation).

The University of Edinburgh's submission to the WMT22 code-mixing shared task (MixMT)

Abstract / Description of output

Publication series

Conference

Access to Document

Fingerprint

Cite this