SHEF-NN: translation quality estimation with neural networks
Abstract
We describe our systems for Tasks 1 and 2 of the WMT15 Shared Task on Quality Estimation. Our submissions use (i) a continuous space language model to extract additional features for Task 1 (SHEFGP, SHEF-SVM), (ii) a continuous bagof-words model to produce word embeddings as features for Task 2 (SHEF-W2V) and (iii) a combination of features produced by QuEst++ and a feature produced with word embedding models (SHEFQuEst++). Our systems outperform the baseline as well as many other submissions. The results are especially encouraging for Task 2, where our best performing system (SHEF-W2V) only uses features learned in an unsupervised fashion.Citation
Shah, K., Logacheva, V., Paetzold, G., Blain, F., Beck, D., Bougares, F. and Specia, L. (2015) SHEF-NN: translation quality estimation with neural networks, Proceedings of the Tenth Workshop on Statistical Machine Translation, 17-18 September, 2015, Lisbon, Portugal.Additional Links
https://www.aclweb.org/anthology/W15-3041/Type
Conference contributionLanguage
enDescription
© 2015 The Authors. Published by Association for Computational Linguistics. This is an open access article available under a Creative Commons licence. The published version can be accessed at the following link on the publisher’s website: https://www.aclweb.org/anthology/W15-3041ISBN
9781941643327Scopus Count
Except where otherwise noted, this item's license is described as https://creativecommons.org/licenses/by-nc-sa/4.0/