Domain Adaptation for Statistical Machine Translation

TL;DR

This paper investigates domain adaptation techniques to enhance statistical machine translation performance across specific fields, addressing challenges like ambiguity, style variation, and OOV words with novel solutions.

Contribution

It explores and proposes effective domain adaptation methods tailored for SMT, focusing on resolving ambiguity, style differences, and OOV issues in domain-specific translation.

Findings

Improved translation quality in domain-specific SMT systems

Effective handling of ambiguity and style variation

Reduction in out-of-vocabulary word problems

Abstract

Statistical machine translation (SMT) systems perform poorly when it is applied to new target domains. Our goal is to explore domain adaptation approaches and techniques for improving the translation quality of domain-specific SMT systems. However, translating texts from a specific domain (e.g., medicine) is full of challenges. The first challenge is ambiguity. Words or phrases contain different meanings in different contexts. The second one is language style due to the fact that texts from different genres are always presented in different syntax, length and structural organization. The third one is the out-of-vocabulary words (OOVs) problem. In-domain training data are often scarce with low terminology coverage. In this thesis, we explore the state-of-the-art domain adaptation approaches and propose effective solutions to address those problems.