Two Approaches to Topic Modeling Within an Encyclopedic Corpus

• Activity Number: 290
• Type: Invited
• Date/Time: Tuesday, August 2, 2016 : 8:30 AM to 10:20 AM
• Sponsor: Section for Statistical Programmers and Analysts
• Author(s): Lauren Tilton*
• Companies: Yale University
• Keywords: topic modelling ; natural language processing ; information retrieval ; document clustering

Abstract

Using the Stanford Encyclopedia of Philosophy, we compare latent Dirichlet allocation and information retrieval for topic detection and document clustering. The focus is on the relative strengths and weaknesses of each for data analysis, rather than a theoretical description of the two techniques. Code to implement both approaches will be provided via an external link.