Document-level Neural Machine Translation with Associated Memory Network

TL;DR

This paper introduces a document-aware memory network to enhance neural machine translation by effectively utilizing detailed document-level context, significantly improving translation quality over standard Transformer models.

Contribution

It proposes a novel memory network architecture that captures rich document context, integrated with Transformer NMT, to improve translation accuracy.

Findings

Significant performance improvements over Transformer baselines.

Effective modeling of detailed document-level context.

Outperforms related document-level NMT approaches.

Abstract

Standard neural machine translation (NMT) is on the assumption that the document-level context is independent. Most existing document-level NMT approaches are satisfied with a smattering sense of global document-level information, while this work focuses on exploiting detailed document-level context in terms of a memory network. The capacity of the memory network that detecting the most relevant part of the current sentence from memory renders a natural solution to model the rich document-level context. In this work, the proposed document-aware memory network is implemented to enhance the Transformer NMT baseline. Experiments on several tasks show that the proposed method significantly improves the NMT performance over strong Transformer baselines and other related studies.