Topic modeling is used to analyze the distribution of words in documents. It assigns sets of words to “topics,” where each document contains one or more topics. Usually the topic assignments for words and documents is done using unsupervised methods, and don't correspond to a particular definition of topics. A popular method for topic modeling is Latent Dirichlet Allocation (LDA).

• Wikipedia - Topic Model
• Blei 2012 - Introduction to Probabilistic Topic Models

• Classic Papers
  • Blei et al 2003 - Latent Dirichlet Allocation
• Recent papers
  • Schofield et al 2017 - Pulling Out the Stops: Rethinking Stopword Removal for Topic Models
  • Cheevaprawatdomrong et al 2021 - More Than Words: Collocation Tokenization for Latent Dirichlet Allocation Models

• David Blei
• Alexandra Schofield