TY - GEN
T1 - A character-level decoder without explicit segmentation for neural machine translation
AU - Chung, Junyoung
AU - Cho, Kyunghyun
AU - Bengio, Yoshua
PY - 2016
Y1 - 2016
N2 - The existing machine translation systems, whether phrase-based or neural, have relied almost exclusively on word-level modelling with explicit segmentation. In this paper, we ask a fundamental question: can neural machine translation generate a character sequence without any explicit segmentation? To answer this question, we evaluate an attention-based encoderdecoder with a subword-level encoder and a character-level decoder on four language pairs-En-Cs, En-De, En-Ru and En-Fiusing the parallel corpora from WMT'15. Our experiments show that the models with a character-level decoder outperform the ones with a subword-level decoder on all of the four language pairs. Furthermore, the ensembles of neural models with a character-level decoder outperform the state-of-the-art non-neural machine translation systems on En-Cs, En-De and En-Fi and perform comparably on En-Ru.
AB - The existing machine translation systems, whether phrase-based or neural, have relied almost exclusively on word-level modelling with explicit segmentation. In this paper, we ask a fundamental question: can neural machine translation generate a character sequence without any explicit segmentation? To answer this question, we evaluate an attention-based encoderdecoder with a subword-level encoder and a character-level decoder on four language pairs-En-Cs, En-De, En-Ru and En-Fiusing the parallel corpora from WMT'15. Our experiments show that the models with a character-level decoder outperform the ones with a subword-level decoder on all of the four language pairs. Furthermore, the ensembles of neural models with a character-level decoder outperform the state-of-the-art non-neural machine translation systems on En-Cs, En-De and En-Fi and perform comparably on En-Ru.
UR - http://www.scopus.com/inward/record.url?scp=85011898454&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85011898454&partnerID=8YFLogxK
U2 - 10.18653/v1/p16-1160
DO - 10.18653/v1/p16-1160
M3 - Conference contribution
AN - SCOPUS:85011898454
T3 - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016 - Long Papers
SP - 1693
EP - 1703
BT - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016 - Long Papers
PB - Association for Computational Linguistics (ACL)
T2 - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016
Y2 - 7 August 2016 through 12 August 2016
ER -