Machine Translation in the Covid domain: an English-Irish case study for LoResMT 2021

TL;DR

It is demonstrated that extending an 8k in-domain baseline dataset by just 5k lines improved the BLEU score by 27 points.

Abstract

Translation models for the specific domain of translating Covid data from English to Irish were developed for the LoResMT 2021 shared task. Domain adaptation techniques, using a Covid-adapted generic 55k corpus from the Directorate General of Translation, were applied. Fine-tuning, mixed fine-tuning and combined dataset approaches were compared with models trained on an extended in-domain dataset. As part of this study, an English-Irish dataset of Covid related data, from the Health and Education domains, was developed. The highest-performing model used a Transformer architecture trained with an extended in-domain Covid dataset. In the context of this study, we have demonstrated that extending an 8k in-domain baseline dataset by just 5k lines improved the BLEU score by 27 points.

Figures (2)

• Figure 1: Proposed Approach. Optimal hyperparameters are applied to Transformer models which are trained using one of several possible approaches. The training dataset composition is determined by the chosen approach. Models are subsequently evaluated using a suite of metrics.
• Figure 2: Translation performance of all approaches using Transformers with 2 heads