Abstract
Word-level auto-completion (WLAC) plays a crucial role in Computer-Assisted Translation. In this paper, we describe the SJTU-MTLAB’s submission to the WMT23 WLAC task. We propose a joint method to incorporate the machine translation task to the WLAC task. The proposed approach is general and can be applied to various encoder-based architectures. Through extensive experiments, we demonstrate that our approach can greatly improve performance, while maintaining significantly small model sizes.- Anthology ID:
- 2023.wmt-1.77
- Volume:
- Proceedings of the Eighth Conference on Machine Translation
- Month:
- December
- Year:
- 2023
- Address:
- Singapore
- Editors:
- Philipp Koehn, Barry Haddow, Tom Kocmi, Christof Monz
- Venue:
- WMT
- SIG:
- SIGMT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 872–876
- Language:
- URL:
- https://aclanthology.org/2023.wmt-1.77
- DOI:
- 10.18653/v1/2023.wmt-1.77
- Cite (ACL):
- Xingyu Chen and Rui Wang. 2023. SJTU-MTLAB’s Submission to the WMT23 Word-Level Auto Completion Task. In Proceedings of the Eighth Conference on Machine Translation, pages 872–876, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- SJTU-MTLAB’s Submission to the WMT23 Word-Level Auto Completion Task (Chen & Wang, WMT 2023)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2023.wmt-1.77.pdf