Applications, Extensions and Similar Techniques
Topic modeling is a classic problem in information retrieval. Related models and techniques are, among others, latent semantic indexing, independent component analysis, probabilistic latent semantic indexing, non-negative matrix factorization, and Gamma-Poisson.
The LDA model is highly modular and can therefore be easily extended. The main field of interest is modeling relations between topics. This is achieved by using another distribution on the simplex instead of the Dirichlet. The Correlated Topic Model follows this approach, inducing a correlation structure between topics by using the logistic normal distribution instead of the Dirichlet. Another extension is the hierarchical LDA (hLDA), where topics are joined together in a hierarchy by using the nested Chinese restaurant process. LDA can also be extended to a corpus in which a document includes two types of information (e.g., words and names), as in the LDA-dual model. Supervised versions of LDA include L-LDA, which has been found to perform better than SVM under certain conditions. Nonparametric extensions of LDA include the Hierarchical Dirichlet process mixture model, which allows the number of topics to be unbounded and learnt from data and the Nested Chinese Restaurant Process which allows topics to be arranged in a hierarchy whose structure is learnt from data.
As noted earlier, PLSA is similar to LDA. The LDA model is essentially the Bayesian version of PLSA model. The Bayesian formulation tends to perform better on small datasets because Bayesian methods can avoid overfitting the data. In a very large dataset, the results are probably the same. One difference is that PLSA uses a variable to represent a document in the training set. So in PLSA, when presented with a document the model hasn't seen before, we fix --the probability of words under topics—to be that learned from the training set and use the same EM algorithm to infer --the topic distribution under . Blei argues that this step is cheating because you are essentially refitting the model to the new data.
Variations on LDA have been used to automatically put natural images into categories, such as "bedroom" or "forest", by treating an image as a document, and small patches of the image as words; one of the variations is called Spatial Latent Dirichlet Allocation.
Read more about this topic: Latent Dirichlet Allocation
Famous quotes containing the words extensions, similar and/or techniques:
“If we focus exclusively on teaching our children to read, write, spell, and count in their first years of life, we turn our homes into extensions of school and turn bringing up a child into an exercise in curriculum development. We should be parents first and teachers of academic skills second.”
—Neil Kurshan (20th century)
“We cannot feel strongly toward the totally unlike because it is unimaginable, unrealizable; nor yet toward the wholly like because it is stale—identity must always be dull company. The power of other natures over us lies in a stimulating difference which causes excitement and opens communication, in ideas similar to our own but not identical, in states of mind attainable but not actual.”
—Charles Horton Cooley (1864–1929)
“It is easy to lose confidence in our natural ability to raise children. The true techniques for raising children are simple: Be with them, play with them, talk to them. You are not squandering their time no matter what the latest child development books say about “purposeful play” and “cognitive learning skills.””
—Neil Kurshan (20th century)