Multi-Domain Neural Machine Translation

TL;DR

This paper introduces a multi-domain neural machine translation model that treats text domains as distinct languages, enabling effective domain switching and achieving superior translation quality over traditional fine-tuning methods.

Contribution

It proposes a novel multi-domain NMT approach using multilingual methods, demonstrating its advantages and exploring domain knowledge requirements.

Findings

Multi-domain NMT outperforms fine-tuning in translation quality.

Treating domains as languages enables effective domain switching.

High translation quality is achievable even without pre-specified domain knowledge.

Abstract

We present an approach to neural machine translation (NMT) that supports multiple domains in a single model and allows switching between the domains when translating. The core idea is to treat text domains as distinct languages and use multilingual NMT methods to create multi-domain translation systems, we show that this approach results in significant translation quality gains over fine-tuning. We also explore whether the knowledge of pre-specified text domains is necessary, turns out that it is after all, but also that when it is not known quite high translation quality can be reached.