Explain Latent Dirichlet Allocation (LDA).

Latent Dirichlet Allocation (LDA) is a common method of topic modeling, or classifying documents by subject matter.

LDA is a generative model that represents documents as a mixture of topics that each have their own probability distribution of possible words.

The “Dirichlet” distribution is simply a distribution of distributions. In LDA, documents are distributions of topics that are distributions of words.