Enhancing Machine Translation with Dependency-Aware Self-Attention
This work addresses translation accuracy issues for users in low-resource languages and long-sentence contexts, representing an incremental improvement over existing methods.
The authors tackled the problem of neural machine translation by incorporating syntactic knowledge into the Transformer model, resulting in improved translation quality, particularly for long sentences and low-resource scenarios, as demonstrated on WMT and WAT datasets.
Most neural machine translation models only rely on pairs of parallel sentences, assuming syntactic information is automatically learned by an attention mechanism. In this work, we investigate different approaches to incorporate syntactic knowledge in the Transformer model and also propose a novel, parameter-free, dependency-aware self-attention mechanism that improves its translation quality, especially for long sentences and in low-resource scenarios. We show the efficacy of each approach on WMT English-German and English-Turkish, and WAT English-Japanese translation tasks.