KenMeSH: Knowledge-enhanced End-to-end Biomedical Text Labelling
This addresses the problem of large-scale biomedical document indexing for researchers and database managers, though it appears incremental as it builds on existing methods with enhancements.
The authors tackled the challenge of automatically assigning Medical Subject Headings (MeSH) to biomedical articles by proposing KenMeSH, an end-to-end model that integrates text features with knowledge-enhanced attention, achieving state-of-the-art performance on multiple measures.
Currently, Medical Subject Headings (MeSH) are manually assigned to every biomedical article published and subsequently recorded in the PubMed database to facilitate retrieving relevant information. With the rapid growth of the PubMed database, large-scale biomedical document indexing becomes increasingly important. MeSH indexing is a challenging task for machine learning, as it needs to assign multiple labels to each article from an extremely large hierachically organized collection. To address this challenge, we propose KenMeSH, an end-to-end model that combines new text features and a dynamic \textbf{K}nowledge-\textbf{en}hanced mask attention that integrates document features with MeSH label hierarchy and journal correlation features to index MeSH terms. Experimental results show the proposed method achieves state-of-the-art performance on a number of measures.