Sparse Additive Generative Model of Text
Matlab Jupyter Notebook C Python M


This is a soft launch for the distribution of the code from our paper,

Sparse Additive Generative Models of Text
Jacob Eisenstein, Amr Ahmed, and Eric P. Xing
International Conference on Machine Learning, 2011

Code (c) 2011 Jacob Eisenstein
Licensed under the Apache License, version 2.0

This distribution includes borrowed code from lots of other smart people. It's all in the directory called "3rd-party":

1) fastfit, lightspeed, util.h, and util.c are from Tom Minka. These are released under the MIT license.
2) lda-eval is a slightly-modified version of the LDA evaluation code from Wallach, Murray, Salakhudinov, and Mimno [MIT license]
3) minimize.m is from Carl Rasmussen 
4) process_options.m is from Mark A. Paskin, under the GPL 2
5) tprod is from Jason Farquhar, under the license found in its directory

See "HOWTO" for how to get started.