FRMT: A Benchmark for Few-Shot Region-Aware Machine Translation

Parker Riley, Timothy Dozat, Jan A. Botha, Xavier Garcia, Dan Garrette, Jason Riesa, Orhan Firat, Noah Constant


Abstract
We present FRMT, a new dataset and evaluation benchmark for Few-shot Region-aware Machine Translation, a type of style-targeted translation. The dataset consists of professional translations from English into two regional variants each of Portuguese and Mandarin Chinese. Source documents are selected to enable detailed analysis of phenomena of interest, including lexically distinct terms and distractor terms. We explore automatic evaluation metrics for FRMT and validate their correlation with expert human evaluation across both region-matched and mismatched rating scenarios. Finally, we present a number of baseline models for this task, and offer guidelines for how researchers can train, evaluate, and compare their own models. Our dataset and evaluation code are publicly available: https://bit.ly/frmt-task.
Anthology ID:
2023.tacl-1.39
Volume:
Transactions of the Association for Computational Linguistics, Volume 11
Month:
Year:
2023
Address:
Cambridge, MA
Venue:
TACL
SIG:
Publisher:
MIT Press
Note:
Pages:
671–685
Language:
URL:
https://aclanthology.org/2023.tacl-1.39
DOI:
10.1162/tacl_a_00568
Bibkey:
Cite (ACL):
Parker Riley, Timothy Dozat, Jan A. Botha, Xavier Garcia, Dan Garrette, Jason Riesa, Orhan Firat, and Noah Constant. 2023. FRMT: A Benchmark for Few-Shot Region-Aware Machine Translation. Transactions of the Association for Computational Linguistics, 11:671–685.
Cite (Informal):
FRMT: A Benchmark for Few-Shot Region-Aware Machine Translation (Riley et al., TACL 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/landing_page/2023.tacl-1.39.pdf
Video:
 https://preview.aclanthology.org/landing_page/2023.tacl-1.39.mp4