Large Scale Translation Quality Estimation

This study explores methods for developing a large scale Quality Estimation framework for Machine Translation. We expand existing resources for Quality Estimation across related languages by using different transfer learning methods. The transfer learning methods are: Transductive SVM, Label Propagation and Self-taught Learning. We use transfer learning methods on the available labelled datasets, e.g. en-es, to produce a range of Quality Estimation models for Romance languages, while also adapting for subtitling as a new domain. The Self-taught Learning method shows the most promising results among the used techniques.