IST-Unbabel Participation in the WMT20 Quality Estimation Shared Task
2020-11-01WMT (EMNLP) 2020Unverified0· sign in to hype
João Moura, Miguel Vera, Daan van Stigt, Fabio Kepler, André F. T. Martins
Unverified — Be the first to reproduce this paper.
ReproduceAbstract
We present the joint contribution of IST and Unbabel to the WMT 2020 Shared Task on Quality Estimation. Our team participated on all tracks (Direct Assessment, Post-Editing Effort, Document-Level), encompassing a total of 14 submissions. Our submitted systems were developed by extending the OpenKiwi framework to a transformer-based predictor-estimator architecture, and to cope with glass-box, uncertainty-based features coming from neural machine translation systems.