BioSentVec: creating sentence embeddings for biomedical texts
This addresses a gap for researchers and developers in biomedical text mining by providing the first open set of sentence embeddings for this domain.
The authors tackled the lack of pre-trained sentence embeddings for biomedical texts by introducing BioSentVec, trained on over 30 million documents from PubMed and MIMIC-III, which achieved state-of-the-art performance in sentence similarity tasks.
Sentence embeddings have become an essential part of today's natural language processing (NLP) systems, especially together advanced deep learning methods. Although pre-trained sentence encoders are available in the general domain, none exists for biomedical texts to date. In this work, we introduce BioSentVec: the first open set of sentence embeddings trained with over 30 million documents from both scholarly articles in PubMed and clinical notes in the MIMIC-III Clinical Database. We evaluate BioSentVec embeddings in two sentence pair similarity tasks in different text genres. Our benchmarking results demonstrate that the BioSentVec embeddings can better capture sentence semantics compared to the other competitive alternatives and achieve state-of-the-art performance in both tasks. We expect BioSentVec to facilitate the research and development in biomedical text mining and to complement the existing resources in biomedical word embeddings. BioSentVec is publicly available at https://github.com/ncbi-nlp/BioSentVec