MeSHSim: An R/Bioconductor package for measuring semantic similarity over MeSH headings and MEDLINE documents

All recent MEDLINE documents are indexed by Medical Subject Headings (MeSH). Computing semantic similarity between two MeSH headings as well as two documents has become very important for many biomedical text mining applications. We develop an R package, MeSHSim, which can compute nine similarity measures between MeSH nodes, by which similarity between MeSH Headings as well as MEDLINE documents can be computed. In addition, MeSHSim supports querying hierarchy information of a MeSH heading and retrieving MeSH headings of a query document. It can be easily integrated into pipelines for any biomedical text analysis tasks. MeSHSim is released under GPL(General Public License), and available through Bioconductor and from Github at https://github.com/JingZhou2015/MeSHSim.