QEMind: Alibaba's Submission to the WMT21 Quality Estimation Shared Task
This work addresses quality control for machine translation users, but it is incremental as it builds on existing pre-trained models and features.
The paper tackles the problem of quality estimation for machine translation without reference translations, presenting QEMind, a system that uses XLM-Roberta and additional features to outperform the best system from the previous year's Direct Assessment task.
Quality Estimation, as a crucial step of quality control for machine translation, has been explored for years. The goal is to investigate automatic methods for estimating the quality of machine translation results without reference translations. In this year's WMT QE shared task, we utilize the large-scale XLM-Roberta pre-trained model and additionally propose several useful features to evaluate the uncertainty of the translations to build our QE system, named \textit{QEMind}. The system has been applied to the sentence-level scoring task of Direct Assessment and the binary score prediction task of Critical Error Detection. In this paper, we present our submissions to the WMT 2021 QE shared task and an extensive set of experimental results have shown us that our multilingual systems outperform the best system in the Direct Assessment QE task of WMT 2020.