Detecting hotspots in insulin-like growth factors 1 research through MetaMap and data mining technologies

Abstract

Most digital information resources for readers of the medical library exist in the form of unstructured free text (journal papers). Therefore it has become the new direction of data mining research to extract keywords in the collection of medical literature and turn them into structured knowledge that is easily accessible and analyzable. MetaMap, a mapping tool from free text to the UMLS Metathesaurus developed by the U.S. National Library of Medicine, maps keywords to the normative UMLS thesaurus, and provides a rating for the mapping degree of every word. The present study extracts keywords from the English language literature of insulin-like growth factors 1 research, assigns weights to the keywords using the BM25F model, screens out groups of important keywords, carries out a cluster analysis of these keywords.