Character-Level Translation with Self-attention

Yingqiang Gao, Nikola I. Nikolov, Yuhuang Hu, Richard H.R. Hahnloser


Abstract
We explore the suitability of self-attention models for character-level neural machine translation. We test the standard transformer model, as well as a novel variant in which the encoder block combines information from nearby characters using convolutions. We perform extensive experiments on WMT and UN datasets, testing both bilingual and multilingual translation to English using up to three input languages (French, Spanish, and Chinese). Our transformer variant consistently outperforms the standard transformer at the character-level and converges faster while learning more robust character-level alignments.
Anthology ID:
2020.acl-main.145
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1591–1604
Language:
URL:
https://www.aclweb.org/anthology/2020.acl-main.145
DOI:
10.18653/v1/2020.acl-main.145
Bib Export formats:
BibTeX MODS XML EndNote
PDF:
http://aclanthology.lst.uni-saarland.de/2020.acl-main.145.pdf
Video:
 http://slideslive.com/38929029