PubMedMiner: Mining and Visualizing MeSH-based Associations in PubMed

AMIA Annu Symp Proc. 2014 Nov 14:2014:1990-9. eCollection 2014.

Abstract

The exponential growth of biomedical literature provides the opportunity to develop approaches for facilitating the identification of possible relationships between biomedical concepts. Indexing by Medical Subject Headings (MeSH) represent high-quality summaries of much of this literature that can be used to support hypothesis generation and knowledge discovery tasks using techniques such as association rule mining. Based on a survey of literature mining tools, a tool implemented using Ruby and R - PubMedMiner - was developed in this study for mining and visualizing MeSH-based associations for a set of MEDLINE articles. To demonstrate PubMedMiner's functionality, a case study was conducted that focused on identifying and comparing comorbidities for asthma in children and adults. Relative to the tools surveyed, the initial results suggest that PubMedMiner provides complementary functionality for summarizing and comparing topics as well as identifying potentially new knowledge.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Data Mining*
  • MEDLINE
  • Medical Subject Headings*
  • PubMed*
  • Unified Medical Language System