Dual Past and Future for Neural Machine Translation
This work addresses translation adequacy issues in NMT, which is crucial for improving machine translation quality, though it appears incremental as it builds on existing Past and Future modeling approaches.
The paper tackles the inadequate-translation problem in Neural Machine Translation by introducing a dual framework that uses source-to-target and target-to-source models to better supervise Past and Future modules, resulting in significant improvements in translation adequacy and outperforming previous methods on two tasks.
Though remarkable successes have been achieved by Neural Machine Translation (NMT) in recent years, it still suffers from the inadequate-translation problem. Previous studies show that explicitly modeling the Past and Future contents of the source sentence is beneficial for translation performance. However, it is not clear whether the commonly used heuristic objective is good enough to guide the Past and Future. In this paper, we present a novel dual framework that leverages both source-to-target and target-to-source NMT models to provide a more direct and accurate supervision signal for the Past and Future modules. Experimental results demonstrate that our proposed method significantly improves the adequacy of NMT predictions and surpasses previous methods in two well-studied translation tasks.