Curriculum Learning for Domain Adaptation in Neural Machine Translation

TL;DR

This paper presents a curriculum learning method for domain adaptation in neural machine translation, improving performance by training on grouped samples ordered by domain similarity.

Contribution

It introduces a simple, adaptable curriculum learning approach that enhances domain-specific neural machine translation without complex modifications.

Findings

Outperforms unadapted models in two domains and language pairs

Easy to implement on existing neural frameworks

Consistent improvements over baseline methods

Abstract

We introduce a curriculum learning approach to adapt generic neural machine translation models to a specific domain. Samples are grouped by their similarities to the domain of interest and each group is fed to the training algorithm with a particular schedule. This approach is simple to implement on top of any neural framework or architecture, and consistently outperforms both unadapted and adapted baselines in experiments with two distinct domains and two language pairs.