Collapsed Gibbs Samplers and Related Utility Functions for LDA-type Models

Description

This package contains functions to read in text corpora, fit LDA-type models to them, and use the fitted models to explore the data and make predictions.

Details

Author(s)

Jonathan Chang (slycoder@gmail.com) Andrew Dai (a.dai@ed.ac.uk)

Special thanks to the following for their reports and comments: Edo Airoldi, Jordan Boyd-Graber, Christopher E. Cramer, James Danowski, Khalid El-Arini, Roger Levy, Solomon Messing, Joerg Reichardt

References

Blei, David M. and Ng, Andrew and Jordan, Michael. Latent Dirichlet allocation. Journal of Machine Learning Research, 2003.

See Also

Functions to fit models: lda.collapsed.gibbs.sampler slda.em mmsb.collapsed.gibbs.sampler nubbi.collapsed.gibbs.sampler rtm.collapsed.gibbs.sampler

Functions to read/create corpora: lexicalize read.documents read.vocab

Functions to manipulate corpora: concatenate.documents filter.words shift.word.indices links.as.edgelist

Functions to compute summary statistics on corpora: word.counts document.lengths

Functions which use the output of fitted models: predictive.distribution top.topic.words top.topic.documents predictive.link.probability

Included data sets: cora poliblog sampson

Examples

## See demos for the following three common use cases:

## Not run: demo(lda)

## Not run: demo(slda)

## Not run: demo(mmsb)

## Not run: demo(rtm)

Results