Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task

Marcel Bollmann; Rahul Aralikatte; Héctor Ricardo Murrieta Bello; Daniel Hershcovich; Miryam de Lhoneux; Anders Søgaard

doi:10.18653/v1/2021.americasnlp-1.28

Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task

Marcel Bollmann, Rahul Aralikatte, Héctor Ricardo Murrieta Bello, Daniel Hershcovich, Miryam de Lhoneux, Anders Søgaard

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › peer review

7 Citationer (Scopus)

18 Downloads (Pure)

Abstract

We evaluated a range of neural machine translation techniques developed specifically for low-resource scenarios. Unsuccessfully. In the end, we submitted two runs: (i) a standard phrase-based model, and (ii) a random babbling baseline using character trigrams. We found that it was surprisingly hard to beat (i), in spite of this model being, in theory, a bad fit for polysynthetic languages; and more interestingly, that (ii) was better than several of the submitted systems, highlighting how difficult low-resource machine translation for polysynthetic languages is.

Originalsprog	Engelsk
Titel	Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021
Redaktører	Manuel Mager, Arturo Oncevay, Annette Rios, Ivan Vladimir Meza Ruiz, Alexis Palmer, Graham Neubig, Katharina Kann
Forlag	Association for Computational Linguistics
Publikationsdato	2021
Sider	248-254
ISBN (Elektronisk)	9781954085442
DOI	https://doi.org/10.18653/v1/2021.americasnlp-1.28
Status	Udgivet - 2021
Begivenhed	1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021 - Virtual, Online Varighed: 11 jun. 2021 → …

Konference

Konference	1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021
By	Virtual, Online
Periode	11/06/2021 → …

Bibliografisk note

Publisher Copyright:
© 2021 Association for Computational Linguistics

Adgang til dokumentet

10.18653/v1/2021.americasnlp-1.28Licens: Andet

FulltextForlagets udgivne version, 193 KBLicens: Andet

Andre filer og links

Link to publication in Scopus

Citationsformater

Bollmann, M., Aralikatte, R., Bello, H. R. M., Hershcovich, D., de Lhoneux, M., & Søgaard, A. (2021). Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task. I M. Mager, A. Oncevay, A. Rios, I. V. M. Ruiz, A. Palmer, G. Neubig, & K. Kann (red.), Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021 (s. 248-254). Association for Computational Linguistics. https://doi.org/10.18653/v1/2021.americasnlp-1.28

Moses and the Character-Based Random Babbling Baseline : CoAStaL at AmericasNLP 2021 Shared Task. / Bollmann, Marcel; Aralikatte, Rahul; Bello, Héctor Ricardo Murrieta; Hershcovich, Daniel; de Lhoneux, Miryam; Søgaard, Anders.

Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021. red. / Manuel Mager; Arturo Oncevay; Annette Rios; Ivan Vladimir Meza Ruiz; Alexis Palmer; Graham Neubig; Katharina Kann. Association for Computational Linguistics, 2021. s. 248-254.

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › peer review

Bollmann, M, Aralikatte, R, Bello, HRM, Hershcovich, D, de Lhoneux, M & Søgaard, A 2021, Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task. i M Mager, A Oncevay, A Rios, IVM Ruiz, A Palmer, G Neubig & K Kann (red), Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021. Association for Computational Linguistics, s. 248-254, 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021, Virtual, Online, 11/06/2021. https://doi.org/10.18653/v1/2021.americasnlp-1.28

Bollmann M, Aralikatte R, Bello HRM, Hershcovich D, de Lhoneux M, Søgaard A. Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task. I Mager M, Oncevay A, Rios A, Ruiz IVM, Palmer A, Neubig G, Kann K, red., Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021. Association for Computational Linguistics. 2021. s. 248-254 https://doi.org/10.18653/v1/2021.americasnlp-1.28

Bollmann, Marcel ; Aralikatte, Rahul ; Bello, Héctor Ricardo Murrieta ; Hershcovich, Daniel ; de Lhoneux, Miryam ; Søgaard, Anders. / Moses and the Character-Based Random Babbling Baseline : CoAStaL at AmericasNLP 2021 Shared Task. Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021. red. / Manuel Mager ; Arturo Oncevay ; Annette Rios ; Ivan Vladimir Meza Ruiz ; Alexis Palmer ; Graham Neubig ; Katharina Kann. Association for Computational Linguistics, 2021. s. 248-254

@inproceedings{b7056869ad7b4215afd2729d0e492ffd,

title = "Moses and the Character-Based Random Babbling Baseline: CoAStaL at AmericasNLP 2021 Shared Task",

abstract = "We evaluated a range of neural machine translation techniques developed specifically for low-resource scenarios. Unsuccessfully. In the end, we submitted two runs: (i) a standard phrase-based model, and (ii) a random babbling baseline using character trigrams. We found that it was surprisingly hard to beat (i), in spite of this model being, in theory, a bad fit for polysynthetic languages; and more interestingly, that (ii) was better than several of the submitted systems, highlighting how difficult low-resource machine translation for polysynthetic languages is.",

author = "Marcel Bollmann and Rahul Aralikatte and Bello, {H{\'e}ctor Ricardo Murrieta} and Daniel Hershcovich and {de Lhoneux}, Miryam and Anders S{\o}gaard",

note = "Publisher Copyright: {\textcopyright} 2021 Association for Computational Linguistics; 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021 ; Conference date: 11-06-2021",

year = "2021",

doi = "10.18653/v1/2021.americasnlp-1.28",

language = "English",

pages = "248--254",

editor = "Manuel Mager and Arturo Oncevay and Annette Rios and Ruiz, {Ivan Vladimir Meza} and Alexis Palmer and Graham Neubig and Katharina Kann",

booktitle = "Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021",

publisher = "Association for Computational Linguistics",

}

TY - GEN

T1 - Moses and the Character-Based Random Babbling Baseline

T2 - 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021

AU - Bollmann, Marcel

AU - Aralikatte, Rahul

AU - Bello, Héctor Ricardo Murrieta

AU - Hershcovich, Daniel

AU - de Lhoneux, Miryam

AU - Søgaard, Anders

PY - 2021

Y1 - 2021

N2 - We evaluated a range of neural machine translation techniques developed specifically for low-resource scenarios. Unsuccessfully. In the end, we submitted two runs: (i) a standard phrase-based model, and (ii) a random babbling baseline using character trigrams. We found that it was surprisingly hard to beat (i), in spite of this model being, in theory, a bad fit for polysynthetic languages; and more interestingly, that (ii) was better than several of the submitted systems, highlighting how difficult low-resource machine translation for polysynthetic languages is.

AB - We evaluated a range of neural machine translation techniques developed specifically for low-resource scenarios. Unsuccessfully. In the end, we submitted two runs: (i) a standard phrase-based model, and (ii) a random babbling baseline using character trigrams. We found that it was surprisingly hard to beat (i), in spite of this model being, in theory, a bad fit for polysynthetic languages; and more interestingly, that (ii) was better than several of the submitted systems, highlighting how difficult low-resource machine translation for polysynthetic languages is.

UR - http://www.scopus.com/inward/record.url?scp=85123952292&partnerID=8YFLogxK

U2 - 10.18653/v1/2021.americasnlp-1.28

DO - 10.18653/v1/2021.americasnlp-1.28

M3 - Article in proceedings

AN - SCOPUS:85123952292

SP - 248

EP - 254

BT - Proceedings of the 1st Workshop on Natural Language Processing for Indigenous Languages of the Americas, AmericasNLP 2021

A2 - Mager, Manuel

A2 - Oncevay, Arturo

A2 - Rios, Annette

A2 - Ruiz, Ivan Vladimir Meza

A2 - Palmer, Alexis

A2 - Neubig, Graham

A2 - Kann, Katharina

PB - Association for Computational Linguistics

Y2 - 11 June 2021

ER -