MIPT System for World-Level Quality Estimation

Abstract

We explore different model architectures for the WMT 19 shared task on word-level quality estimation of automatic translation. We start with a model similar to Shef-bRNN (Ive et al., 2018), which we modify by using conditional random fields (CRFs) (Lafferty et al., 2001) for sequence labelling. Additionally, we use a different approach for labelling gaps and source words. We further develop this model by including features from different sources such as BERT (Devlin et al., 2018), baseline features for the task (Specia et al., 2018) and transformer encoders (Vaswani et al., 2017). We evaluate the performance of our models on the English-German dataset for the corresponding task.