NLM at BioASQ Synergy 2021: Deep learning-based methods for biomedical semantic question answering about COVID-19
Mourad Sarrouti, Deepak Kumar Gupta, Asma Ben Abacha, Dina Demner‐Fushman · Cross-Language Evaluation Forum · 2021
The COVID-19 outbreak has heightened the need for systems that enable information seekers to search vast corpora of scientific articles to find answers to their natural language questions. This paper describes the participation of the U.S. National Library of Medicine (NLM) team in BioASQ Task Synergy on biomedical semantic question answering for COVID-19. In this work, we exploited the pre-trained Transformer models such as T5 and BART for document re-ranking, passage retrieval, and answer generation. Official results show that among the participating systems, our models achieve strong performance in document retrieval, passage retrieval, and the “ideal answer” generation task. © 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).