Lexically Cohesive Neural Machine Translation with Copy Mechanism
This work addresses lexical cohesion for document-level translation, offering an incremental improvement over existing methods.
The paper tackled the problem of maintaining consistent word choices in document-level neural machine translation by integrating a copy mechanism to reuse words from previous translations, resulting in significant improvements in lexical cohesion compared to prior context-aware models.
Lexically cohesive translations preserve consistency in word choices in document-level translation. We employ a copy mechanism into a context-aware neural machine translation model to allow copying words from previous translation outputs. Different from previous context-aware neural machine translation models that handle all the discourse phenomena implicitly, our model explicitly addresses the lexical cohesion problem by boosting the probabilities to output words consistently. We conduct experiments on Japanese to English translation using an evaluation dataset for discourse translation. The results showed that the proposed model significantly improved lexical cohesion compared to previous context-aware models.