Easy function for text-mining the 'PubMed' repository based on defined sets of terms. The relationship between fix-terms (related to your research topic) and pub-terms (terms which pivot around your research focus) is calculated using the pointwise mutual information algorithm ('PMI'). Church, Kenneth Ward and Hanks, Patrick (1990) < https://www.aclweb.org/anthology/J90-1003/> A text file is generated with the 'PMI'-scores for each fix-term. Then for each collocation pairs (a fix-term + a pub-term), a text file is generated with related article titles and publishing years. Additional Author section will follow in the next version updates.
Term-based text-Mining of the 'PubMed' repository. Requires internet access!
The function takes vectors of string as term sets for fix-terms (research topic focus) and pub-terms (terms which can pivot around your research focus).
Install the function with devtools::install_github("Moshkante/pubmed_mining") or simply via CRAN.
Mining results are given in text format with pointwise mutual information ('PMI'-score) and related article titles and publishing years.
More features will be added in the next version.
Example:
fixterms = "bike" #or multiple terms pubterms = c("dangerous", "extreme") output = getwd() #or "YOUR/DESIRED/PATHWAY" pubmed_textmining(fixterms, pubterms, output)