Dynamically Composing Domain-Data Selection with Clean-Data Selection by “Co-Curricular Learning” for Neural Machine Translation

Wei Wang, Isaac Caswell, Ciprian Chelba


Abstract
Noise and domain are important aspects of data quality for neural machine translation. Existing research focus separately on domain-data selection, clean-data selection, or their static combination, leaving the dynamic interaction across them not explicitly examined. This paper introduces a “co-curricular learning” method to compose dynamic domain-data selection with dynamic clean-data selection, for transfer learning across both capabilities. We apply an EM-style optimization procedure to further refine the “co-curriculum”. Experiment results and analysis with two domains demonstrate the effectiveness of the method and the properties of data scheduled by the co-curriculum.
Anthology ID:
P19-1123
Volume:
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2019
Address:
Florence, Italy
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1282–1292
Language:
URL:
https://www.aclweb.org/anthology/P19-1123
DOI:
10.18653/v1/P19-1123
Bib Export formats:
BibTeX MODS XML EndNote
PDF:
http://aclanthology.lst.uni-saarland.de/P19-1123.pdf