Learning Bilingual Word Representations by Marginalizing Alignments
Computation and Language
2014-05-06 v1
Abstract
We present a probabilistic model that simultaneously learns alignments and distributed representations for bilingual data. By marginalizing over word alignments the model captures a larger semantic context than prior work relying on hard alignments. The advantage of this approach is demonstrated in a cross-lingual classification task, where we outperform the prior published state of the art.
Cite
@article{arxiv.1405.0947,
title = {Learning Bilingual Word Representations by Marginalizing Alignments},
author = {Tomáš Kočiský and Karl Moritz Hermann and Phil Blunsom},
journal= {arXiv preprint arXiv:1405.0947},
year = {2014}
}
Comments
Proceedings of ACL 2014 (Short Papers)