Abstract
Standard neural machine translation (NMT) is on the assumption ofdocument-level context independent. Most existing document-level NMT methodsonly focus on briefly introducing document-level information but fail toconcern about selecting the most related part inside document context. Thecapacity of memory network for detecting the most relevant part of the currentsentence from the memory provides a natural solution for the requirement ofmodeling document-level context by document-level NMT. In this work, we proposea Transformer NMT system with associated memory network (AMN) to both capturethe document-level context and select the most salient part related to theconcerned translation from the memory. Experiments on several tasks show thatthe proposed method significantly improves the NMT performance over strongTransformer baselines and other related studies.