Transformer-based Automatic Post-Editing with a Context-Aware Encoding Approach for Multi-Source Inputs
This work addresses the need for better post-editing in machine translation, though it appears incremental as it builds on existing multi-source APE approaches.
The paper tackles the problem of Automatic Post-Editing (APE) by proposing a Transformer-based model that incorporates source context into machine translation representations, achieving significant improvements over baseline and state-of-the-art multi-source APE models.
Recent approaches to the Automatic Post-Editing (APE) research have shown that better results are obtained by multi-source models, which jointly encode both source (src) and machine translation output (mt) to produce post-edited sentence (pe). Along this trend, we present a new multi-source APE model based on the Transformer. To construct effective joint representations, our model internally learns to incorporate src context into mt representation. With this approach, we achieve a significant improvement over baseline systems, as well as the state-of-the-art multi-source APE model. Moreover, to demonstrate the capability of our model to incorporate src context, we show that the word alignment of the unknown MT system is successfully captured in our encoding results.