Understanding Back-Translation at Scale

Sergey Edunov, Myle Ott, Michael Auli, David Grangier


Abstract
An effective method to improve neural machine translation with monolingual data is to augment the parallel training corpus with back-translations of target language sentences. This work broadens the understanding of back-translation and investigates a number of methods to generate synthetic source sentences. We find that in all but resource poor settings back-translations obtained via sampling or noised beam outputs are most effective. Our analysis shows that sampling or noisy synthetic data gives a much stronger training signal than data generated by beam or greedy search. We also compare how synthetic data compares to genuine bitext and study various domain effects. Finally, we scale to hundreds of millions of monolingual sentences and achieve a new state of the art of 35 BLEU on the WMT’14 English-German test set.
Anthology ID:
D18-1045
Volume:
Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
Month:
October-November
Year:
2018
Address:
Brussels, Belgium
Venue:
EMNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
489–500
Language:
URL:
https://aclanthology.org/D18-1045
DOI:
10.18653/v1/D18-1045
Bibkey:
Cite (ACL):
Sergey Edunov, Myle Ott, Michael Auli, and David Grangier. 2018. Understanding Back-Translation at Scale. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 489–500, Brussels, Belgium. Association for Computational Linguistics.
Cite (Informal):
Understanding Back-Translation at Scale (Edunov et al., EMNLP 2018)
Copy Citation:
PDF:
https://aclanthology.org/D18-1045.pdf
Code
 pytorch/fairseq +  additional community code
Data
EuroparlWMT 2014