A Multilingual Neural Machine Translation Model for Biomedical Data

Abstract

We release a multilingual neural machine translation model, which can be used to translate text in the biomedical domain. The model can translate from 5 languages (French, German, Italian, Korean and Spanish) into English. It is trained with large amounts of generic and biomedical data, using domain tags. Our benchmarks show that it performs near stateof-the-art both on news (generic domain) and biomedical test sets, and that it outperforms the existing publicly released models. We believe that this release will help the large-scale multilingual analysis of the digital content of the COVID-19 crisis and of its effects on society, economy, and healthcare policies. We also release a test set of biomedical text for Korean-English. It consists of 758 sentences from official guidelines and recent papers, all about COVID-19.

The Role of Natural Language Processing during the COVID-19 Pandemic: Health Applications, Opportunities, and Challenges

“…NLP can play an important role to translate these findings and guidelines. For instance, NLP models were trained to offer multilingual translation support for general and biomedical domains [ 94 ]. A separate study constructed a multilingual dataset and then developed a model for cross-lingual intent detection to improve COVID-19 chatbots across the English, Spanish, French, and Spanglish languages [ 95 ].…”