Agreement-based Learning of Parallel Lexicons and Phrases from Non-Parallel Corpora
Computation and Language
2016-06-16 v1
Abstract
We introduce an agreement-based approach to learning parallel lexicons and phrases from non-parallel corpora. The basic idea is to encourage two asymmetric latent-variable translation models (i.e., source-to-target and target-to-source) to agree on identifying latent phrase and word alignments. The agreement is defined at both word and phrase levels. We develop a Viterbi EM algorithm for jointly training the two unidirectional models efficiently. Experiments on the Chinese-English dataset show that agreement-based learning significantly improves both alignment and translation performance.
Cite
@article{arxiv.1606.04597,
title = {Agreement-based Learning of Parallel Lexicons and Phrases from Non-Parallel Corpora},
author = {Chunyang Liu and Yang Liu and Huanbo Luan and Maosong Sun and Heng Yu},
journal= {arXiv preprint arXiv:1606.04597},
year = {2016}
}
Comments
Accepted for publication in the Proceedings of ACL 2016