MeSHSim

MeSHSim computes semantic similarity between Medical Subject Headings (MeSH) and MEDLINE documents in R, providing nine distinct similarity measures to quantify semantic relatedness among MeSH headings and across MEDLINE records for biomedical text mining applications.

Key Features:

• Semantic Similarity Computation: Computes semantic similarity between MeSH nodes and between MeSH-annotated MEDLINE documents using multiple metrics.
• Nine Similarity Measures: Implements nine distinct similarity measures to provide alternative metrics for semantic relatedness.
• Hierarchy Information Querying: Queries the hierarchical structure of MeSH headings to inform similarity calculations based on MeSH taxonomy relationships.
• Retrieval of MeSH Headings: Retrieves relevant MeSH headings from query documents to enable document-level semantic analysis.

Scientific Applications:

• Literature mining: Identifies semantically related MeSH terms and MEDLINE records to support extraction of related biomedical literature.
• Information retrieval: Ranks or filters MEDLINE documents by MeSH-based semantic similarity to improve retrieval relevance.
• Knowledge discovery: Reveals semantic relationships and patterns among medical concepts represented by MeSH headings across MEDLINE.

Methodology:

Leverages the structured indexing of MEDLINE documents by MeSH headings to compute similarity scores and supports nine distinct similarity measures for those computations.