Detecting Topic Drift with Compound Topic Models


  • Dan Knights University of Colorado at Boulder
  • Michael Mozer University of Colorado at Boulder
  • Nicolas Nicolov J. D. Power and Associates


Topic models, Topic tracking, Topic drift, Trend identification and tracking


The Latent Dirichlet Allocation topic model of Blei, Ng, and Jordan (2003) is well-established as an effective approach to recovering meaningful topics of conversation from a set of documents. However, a useful analysis of user-generated content is concerned not only with the recovery of topics from a static data set, but with the evolution of topics over time. We employ a compound topic model (CTM) to track topics across two distinct data sets (i.e. past and present) and to visualize trends in topics over time; we evaluate several metrics for detecting a change in the distribution of topics within a time-window; and we illustrate how our approach discovers emerging conversation topics related to current events in real data sets.




How to Cite

Knights, D., Mozer, M., & Nicolov, N. (2009). Detecting Topic Drift with Compound Topic Models. Proceedings of the International AAAI Conference on Web and Social Media, 3(1), 242-245. Retrieved from