An Effective Coverage Approach for Attention-based Neural Machine Translation

Neural Machine Translation recently has become the state-of-the-art approach in Machine Translation. One of the more advanced techniques concerning this approach, the attention model, tends to not use alignments from past translation steps and selects the context word purely using the devised attention score. Unfortunately, this sometimes leads to repetition and omission of important words in translations. To solve this problem, we propose a simple approach using coverage techniques that can be used in conjunction with a diverse number of attention models. Our experiments show that our improved technique increases the quality of translation on both English Vietnamese and Japanese - Vietnamese language pairings.