SMDT: Selective Memory-Augmented Neural Document Translation
This work addresses the challenge of handling large hypothesis spaces in document translation for machine translation researchers, representing an incremental advancement in context utilization.
The paper tackles the problem of limited context diversity in document-level neural machine translation by proposing a selective memory-augmented model that retrieves similar bilingual sentence pairs to augment global context and uses a selective mechanism to capture local and diverse global contexts, resulting in significant performance improvements over previous models on three public datasets.
Existing document-level neural machine translation (NMT) models have sufficiently explored different context settings to provide guidance for target generation. However, little attention is paid to inaugurate more diverse context for abundant context information. In this paper, we propose a Selective Memory-augmented Neural Document Translation model to deal with documents containing large hypothesis space of the context. Specifically, we retrieve similar bilingual sentence pairs from the training corpus to augment global context and then extend the two-stream attention model with selective mechanism to capture local context and diverse global contexts. This unified approach allows our model to be trained elegantly on three publicly document-level machine translation datasets and significantly outperforms previous document-level NMT models.