Document Graph for Neural Machine Translation
arXiv:2012.03477
Abstract
Previous works have shown that contextual information can improve the performance of neural machine translation (NMT). However, most existing document-level NMT methods only consider a few number of previous sentences. How to make use of the whole document as global contexts is still a challenge. To address this issue, we hypothesize that a document can be represented as a graph that connects relevant contexts regardless of their distances. We employ several types of relations, including adjacency, syntactic dependency, lexical consistency, and coreference, to construct the document graph. Then, we incorporate both source and target graphs into the conventional Transformer architecture with graph convolutional networks. Experiments on various NMT benchmarks, including IWSLT English--French, Chinese-English, WMT English--German and Opensubtitle English--Russian, demonstrate that using document graphs can significantly improve the translation quality. Extensive analysis verifies that the document graph is beneficial for capturing discourse phenomena.
arXiv commentAccepted by EMNLP2021
References in corpus (6)
- Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
- Effective Approaches to Attention-based Neural Machine Translation
- Layer Normalization
- Does Neural Machine Translation Benefit from Larger Context?
- Modeling Coherence for Neural Machine Translation with Dynamic and Topic Caches
- Using Whole Document Context in Neural Machine Translation