Multi-source transformer with combined losses for automatic post editing

Amirhossein Tebbifakhr, Ruchit Agrawal, Matteo Negri, Marco Turchi


Abstract
Recent approaches to the Automatic Post-editing (APE) of Machine Translation (MT) have shown that best results are obtained by neural multi-source models that correct the raw MT output by also considering information from the corresponding source sentence. To this aim, we present for the first time a neural multi-source APE model based on the Transformer architecture. Moreover, we employ sequence-level loss functions in order to avoid exposure bias during training and to be consistent with the automatic evaluation metrics used for the task. These are the main features of our submissions to the WMT 2018 APE shared task, where we participated both in the PBSMT subtask (i.e. the correction of MT outputs from a phrase-based system) and in the NMT subtask (i.e. the correction of neural outputs). In the first subtask, our system improves over the baseline up to -5.3 TER and +8.23 BLEU points ranking second out of 11 submitted runs. In the second one, characterized by the higher quality of the initial translations, we report lower but statistically significant gains (up to -0.38 TER and +0.8 BLEU), ranking first out of 10 submissions.
Anthology ID:
W18-6471
Volume:
Proceedings of the Third Conference on Machine Translation: Shared Task Papers
Month:
October
Year:
2018
Address:
Belgium, Brussels
Venues:
EMNLP | WMT | WS
SIG:
SIGMT
Publisher:
Association for Computational Linguistics
Note:
Pages:
846–852
Language:
URL:
https://www.aclweb.org/anthology/W18-6471
DOI:
10.18653/v1/W18-6471
Bib Export formats:
BibTeX MODS XML EndNote
PDF:
http://aclanthology.lst.uni-saarland.de/W18-6471.pdf