Factorial LDA Code

This software includes an implementation of Factorial LDA (f-LDA), introduced and described in the papers below. See the included README for open licensing information as well as usage instructions and input/output formatting guidelines.

Factorial LDA is a multi-dimensional topic model because tokens are associated with an entire vector of latent variables rather than a single topic variable. This software is closed related to my multi-faceted topic models code, which includes two-dimensional models, the topic-aspect model and cross-collection LDA. I would like to eventually put all of these together, but there are enough differences that I am keeping them separate until I can do more work on it.

• 8/18/2013 - v0.1 - First release.

• Michael J. Paul and Mark Dredze. Factorial LDA: Sparse Multi-Dimensional Models of Text. Advances in Neural Information Processing Systems (NIPS 2012), Lake Tahoe, Nevada. December 2012.
  
  
• Michael J. Paul and Mark Dredze. Drug Extraction from the Web: Summarizing Drug Experiences with Multi-Dimensional Topic Models. To appear at the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2013), Atlanta. June 2013.