Divide and Rule: Effective Pre-Training for Context-Aware Multi-Encoder Translation Models
This work addresses a training bottleneck for context-aware machine translation systems, offering an incremental but effective method to enhance translation quality.
The paper tackles the difficulty of training contextual parameters in multi-encoder translation models due to sparse training signals, by proposing pre-training over split sentence pairs to increase contextual cues and ease context retrieval. Results show consistent improvements in learning contextual parameters across low and high resource settings, as evaluated with BLEU and contrastive test sets.
Multi-encoder models are a broad family of context-aware neural machine translation systems that aim to improve translation quality by encoding document-level contextual information alongside the current sentence. The context encoding is undertaken by contextual parameters, trained on document-level data. In this work, we discuss the difficulty of training these parameters effectively, due to the sparsity of the words in need of context (i.e., the training signal), and their relevant context. We propose to pre-train the contextual parameters over split sentence pairs, which makes an efficient use of the available data for two reasons. Firstly, it increases the contextual training signal by breaking intra-sentential syntactic relations, and thus pushing the model to search the context for disambiguating clues more frequently. Secondly, it eases the retrieval of relevant context, since context segments become shorter. We propose four different splitting methods, and evaluate our approach with BLEU and contrastive test sets. Results show that it consistently improves learning of contextual parameters, both in low and high resource settings.