An Approach Using Style Classification Features for Quality Estimation

Erwan Moreau (1) , Raphaël Rubino (2)

1 School of Computer Science and Statistics [Dublin]
2 NCLT - National Centre for Language Technology

In this paper we describe our participation to the WMT13 Shared Task on Quality Estimation. The main originality of our approach is to include features originally designed to classify text according to some author's style. This implies the use of reference categories, which are meant to represent the quality of the MT output.

Mots clés: Quality Estimation for Machine Translation

Domaines: Traitement du texte et du document

Erwan Moreau, Raphaël Rubino. An Approach Using Style Classification Features for Quality Estimation. Workshop on Statistical Machine Translation (WMT 2013), Aug 2013, Sofia, Bulgaria. pp 429-434. ⟨hal-00860669⟩