Navigating the Local Modes of Big Data: The Case of Topic Models

Margaret E. Roberts, Brandon M. Stewart, Dustin Tingley · 2016

In Computational Social Science: Discovery and Prediction, Cambridge University Press, 2016.

Abstract

This chapter examines multimodality in topic models and related latent variable models for large-scale text analysis. Using structural topic models and a corpus of political blog posts from the 2008 U.S. presidential election, it explains why nonconvex optimization can make topic-model results sensitive to starting values. The chapter shows how these local modes arise, how they can affect substantive interpretation, and how researchers can diagnose and manage sensitivity when using topic models to analyze large, unstructured text collections.

Related Publications

[Paper] stm: An R Package for Structural Topic Models
[Paper] What Makes Foreign Policy Teams Tick: Explaining Variation in Group Performance at Geopolitical Forecasting
[Paper] Computer-Assisted Reading and Discovery for Student Generated Text in Massive Open Online Courses
[Paper] Computer-Assisted Text Analysis for Comparative Politics
[Conference Proceedings] TopicCheck: Interactive Alignment for Assessing Topic Model Stability
[Conference Proceedings] Computer-Assisted Content Analysis: Topic Models for Exploring Multiple Subjective Interpretations
[Paper] Structural Topic Models for Open-Ended Survey Responses
[Other] The Structural Topic Model and Applied Social Science