May 2017 Notes on Text mining and Data Science

The May 2017 meetup focused on R packages for text mining and the types of analysis possible.

R Packages

The tm and tidytext packages were discussed. Throughout the discussion, tidytext was seen as the cleanest way to work with data.frames. The RTextTools package is another package to perform text mining and sentiment analysis.

• tidytext: Text Mining using 'dplyr', 'ggplot2', and Other Tidy Tools CRAN link
• tm: Text Mining Package CRAN link
• RTextTools: Automatic Text Classification via Supervised Learning no longer developed CRAN link

tidytext Tutorials

The following tidytext tutorials are an excellent resource for learning the procedure and protocol for cleaning and analyzing text. The following vignettes and tutorials were the basis for our discussion.

• tidytext vignettes
  • Introduction to tidytext link
  • Tidy Term Frequency and Inverse Document Frequency (tf-idf) link
  • Converting to and from Document-Term Matrices and Corpus objects link
  • Tidy Topic Modeling link
• UC Business Analytics R Programming Guide tidytext tutorials
  • Tidying Text & Word Frequency link
  • Sentiment Analysis link
  • Term vs. Document Frequency link
  • Word Relationships link
  • Converting Between Tidy and Non-tidy Formats link

back