Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Division
Browse By Session Type
Search Tips
Personal Schedule
Sign In
The rise of digital communication technologies poses a promising challenge for communication research. There is an unprecedented amount of communication data ready for the taking, but traditional manual content analysis is prohibitively expensive for very large data sets. By mastering the right tools to analyze this data we can address old and new research problems from a powerful new angle. One of the tools for this task is topic modeling, which is a computational technique for automatically structuring a large corpus of texts based on its latent (i.e. unobserved) topical structure.
With this paper, we contribute to the constructive use of topic modeling in communication research. We provide an overview of the possibilities and challenges in topic modeling, and propose a preliminary set of best practices and validation routines for using topic models in communication research.
Wouter van Atteveldt
Kasper Welbers, U of Leuven
Lubos Steskal, U of Bergen
Carina Jacobi, Ministry of Internal Affairs
Nel Ruigrok, LJS Media Research