Key points are not available for this paper at this time.
Generally, Neural Machine Translation models generate target words in a left-to-right (L2R) manner and fail to exploit any future (right) semantics information, which usually produces an unbalanced translation. Recent works attempt to utilize the right-to-left (R2L) decoder in bidirectional decoding to alleviate this problem. In this paper, we propose a novel Dynamic Interaction Module (DIM) to dynamically exploit target semantics from R2L translation for enhancing the L2R translation quality. Different from other bidirectional decoding approaches, DIM firstly extracts helpful target information through addressing and reading operations, then updates target semantics for tracking the interactive history. Additionally, we further introduce an agreement regularization term into the training objective to narrow the gap between L2R and R2L translations. Experimental results on NIST ChineseEnglish and WMT'16 EnglishRomanian translation tasks show that our system achieves significant improvements over baseline systems, which also reaches comparable results compared to the state-of-the-art Transformer model with much fewer parameters of it.
Shan et al. (Tue,) studied this question.