Multi-Domain Neural Machine Translation
Name
Sander Tars
Abstract
In this thesis 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 propose approach of unsupervised domain assignment and 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, and even higher than with known domains in some cases.
Additionally, we explore the possibility of intra-language style adaptation through zero shot translation. We show that this approach is able to style
adapt, however, with unresolved text deterioration issues.
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 propose approach of unsupervised domain assignment and 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, and even higher than with known domains in some cases.
Additionally, we explore the possibility of intra-language style adaptation through zero shot translation. We show that this approach is able to style
adapt, however, with unresolved text deterioration issues.
Graduation Thesis language
English
Graduation Thesis type
Master - Computer Science
Supervisor(s)
Mark FiĊĦel, PhD
Defence year
2018