Pre-Trained Multilingual Sequence-to-Sequence Models: A Hope for Low-Resource Language Translation?
En-Shiun Annie Lee, Sarubi Thillainathan, Shravan Nayak, Surangika Ranathunga, David Ifeoluwa Adelani, Ruisi Su, Arya D. McCarthy
Abstract
What can pre-trained multilingual sequence-to-sequence models like mBART contribute to translating low-resource languages? We conduct a thorough empirical experiment in 10 languages to ascertain this, considering five factors: (1) the amount of fine-tuning data, (2) the noise in the fine-tuning data, (3) the amount of pre-training data in the model, (4) the impact of domain mismatch, and (5) language typology. In addition to yielding several heuristics, the experiments form a framework for evaluating the data sensitivities of machine translation systems. While mBART is robust to domain differences, its translations for unseen and typologically distant languages remain below 3.0 BLEU. In answer to our title’s question, mBART is not a low-resource panacea; we therefore encourage shifting the emphasis from new models to new data.
BibTeX
@inproceedings{lee-etal-2022-pre,
title = "Pre-Trained Multilingual Sequence-to-Sequence Models: A Hope for Low-Resource Language Translation?",
author = "Lee, En-Shiun Annie and
Thillainathan, Sarubi and
Nayak, Shravan and
Ranathunga, Surangika and
Adelani, David Ifeoluwa and
Su, Ruisi and
McCarthy, Arya D.",
editor = "Muresan, Smaranda and
Nakov, Preslav and
Villavicencio, Aline",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2022",
month = may,
year = "2022",
address = "Dublin, Ireland",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.findings-acl.6/",
doi = "10.18653/v1/2022.findings-acl.6",
pages = "58--67"
}