topicmodels (version 0.1-5)

LDA: Latent Dirichlet Allocation

Description

Estimate a LDA model using for example the VEM algorithm or Gibbs Sampling.

Usage

LDA(x, k, method = "VEM", control = NULL, model = NULL, ...)

Arguments

x

Object of class "DocumentTermMatrix" with term-frequency weighting or an object coercible to a "simple_triplet_matrix" with integer entries.

k

Integer; number of topics.

method

The method to be used for fitting; currently method = "VEM" or method= "Gibbs" are supported.

control

A named list of the control parameters for estimation or an object of class "LDAcontrol".

model

Object of class "LDA" for initialization.

...

Optional arguments. Currently not used.

Value

• LDA() returns an object of class "LDA".

Details

The C code for LDA from David M. Blei and co-authors is used to estimate and fit a latent dirichlet allocation model with the VEM algorithm. For Gibbs Sampling the C++ code from Xuan-Hieu Phan and co-authors is used.

References

Blei D.M., Ng A.Y., Jordan M.I. (2003). Latent Dirichlet Allocation. Journal of Machine Learning Research, 3, 993--1022.

Phan X.H., Nguyen L.M., Horguchi S. (2008). Learning to Classify Short and Sparse Text & Web with Hidden Topics from Large-scale Data Collections. In Proceedings of the 17th International World Wide Web Conference (WWW 2008), pages 91--100, Beijing, China.

See Also

"LDAcontrol"

Examples

data("AssociatedPress", package = "topicmodels")
lda <- LDA(AssociatedPress[1:20,], control = list(alpha = 0.1), k = 2)
lda_inf <- posterior(lda, AssociatedPress[21:30,])