Our work on exploring model uncertainty methods for glass-box Quality Estimation has been accepted for publication at TACL!

Our new paper on using inductive Transfer-Learning for Quality Estimation has been accepted at EAMT and will be published later this year!

Our new paper exploring both text and visual modalities has been accepted for publication at ACL!

We wrote a new paper to report the findings of the last edition of the shared task on Quality Estimation, organised on the occasion of the third Conference on Machine Translation (WMT'18).

Our recent work on designing solutions for quality estimation of machine translation, and in particular for neural-based machine translation, led us to develop and release a new open-source framework.

Our new paper on the combination of Automatic Post-Editing and Quality Estimation for Machine Translation, joint work with the MT Group at FBK, is now published in the proceedings of the Association for Machine Translation in America (AMTA) 2018.

1000 Machine Translation Errors manually annotated at phrase-level over 400 French-English source sentences and their machine translations, along their human post-edited version, and original references.