Topic Models for Text Data: Quick Pointers to Various Resources

• Note that on Dave Blei's page on topic modeling resources is much more up to date than this one! See also Dave's topic modeling mailing list.

Topic Modeling Softare

• MATLAB software. A comprehensive MATLAB software package from Mark Steyvers and Tom Griffiths, using collapsed Gibbs sampling.
• David Blei's Latent Dirichlet Allocation (LDA) code in C, using variational learning.
• Andrew McCallum and his group have implemented topic modeling as part of the more general MALLET toolkit for statistical text analysis.
• Jonathan Chang's R package for topic models, using collapsed Gibbs sampling.
• Alex Smola and Shravan Narayanmurthy's very fast LDA implementation on Hadoop that can "quite comfortably handle a billion documents" (!) according to Alex.

Introductory Papers on Topic Models

• Probabilistic Topic Models, Steyvers and Griffiths (2006). A review paper on topic modeling that is a useful starting point for covering the basic concepts, with illustrative examples.
• Topic Models, Blei and Lafferty (2009). Another very useful starting point for topic modeling.

Examples of Web Browsers built using Topic Models

• [currently down] A topic-based browser of UCI and UCSD faculty research interests built by Dave Newman and Arthur Asuncion at UCI in 2005.
• [currently down] A topic-based browser for 330,000 New York Times articles, by Dave Newman, UCI.
• Wray Buntine's topic-based search interface to Wikipedia.
• Dave Blei and John Lafferty's browsable 100-topic model of journal articles from Science.
• Sean Gerrish's JSTOR discipline browser, built using topic models.