Multi-Domain Neural Machine Translation
This addresses the challenge of domain adaptation in machine translation for users needing flexible, high-quality translations across varied text types, though it is incremental in building on existing multilingual NMT techniques.
The paper tackles the problem of neural machine translation across multiple domains by treating domains as distinct languages and using multilingual methods, resulting in significant translation quality gains over fine-tuning, with high quality achievable even when domain knowledge is not fully known.
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.