Exploiting Semantics in Neural Machine Translation with Graph Convolutional Networks

TL;DR

This paper introduces a novel approach that incorporates semantic-role representations into neural machine translation using Graph Convolutional Networks, leading to improved translation quality.

Contribution

It is the first to integrate predicate-argument structure via GCNs into NMT, enhancing semantic understanding and translation performance.

Findings

BLEU scores improved over baseline models

Semantic-role information enhances translation accuracy

GCNs effectively encode predicate-argument structures

Abstract

Semantic representations have long been argued as potentially useful for enforcing meaning preservation and improving generalization performance of machine translation methods. In this work, we are the first to incorporate information about predicate-argument structure of source sentences (namely, semantic-role representations) into neural machine translation. We use Graph Convolutional Networks (GCNs) to inject a semantic bias into sentence encoders and achieve improvements in BLEU scores over the linguistic-agnostic and syntax-aware versions on the English--German language pair.