Combining Quality Prediction and System Selection for Improved Automatic Translation Output
Radu Soricut, Sushant Narsale · 2012
This paper presents techniques for referencefree, automatic prediction of Machine Translation output quality at both sentence- and document-level. In addition to helping with document-level quality estimation, sentencelevel predictions are used for system selection, improving the quality of the output translations. We present three system selection techniques and perform evaluations that quantify the gains across multiple domains and language pairs. 1