Topic Models: Latent Dirichlet Allocation

This lecture covers the basics of topic modeling, focusing on Latent Dirichlet Allocation (LDA). It explains the process of clustering documents into topics, estimating topic-word distributions, and inferring document-topic distributions. The lecture also discusses the generative process of LDA, the learning algorithms involved, and the limitations of the model. Additionally, it explores the use of LDA in digital humanities, its evaluation metrics, and its application as a form of dimensionality reduction. The lecture concludes with an overview of approximate inference methods and extensions to the LDA model.