Structural Topic Models for the Social Sciences
Date/Time: March 4th, 2:00 pm
Location: 510 Ross Hall
Abstract:
Topic models are widely used natural language processing techniques for the analysis of identifying the topics of texts. The Structural Topic Model (STM) is a general framework for topic modeling with document-level covariate information. The covariates can improve inference and qualitative interpretability and are allowed to affect topical prevalence, topical content or both. In this talk, Dr. Roberts will provide a description and discussion of the STM and present several social science applications.
Bio:
Dr. Roberts is a 2014 PhD from Harvard. Her work has been published in Science, Journal of the American Statistical Association, and several of the leading Political Science journals. Her research interests lie in the intersection of political methodology and the politics of information, with a specific focus on methods of automated content analysis and the politics of censorship in China.
Funded By:
The College of Liberal Arts and Sciences Signature Research Initiative