Probabilistic theories of semantic category formation in cognitive psychology can be used to model how semantic sense may be built up in the aggregate by a network of cooperating users. A model based on Latent Dirichlet Allocation is applied to word activity data in the Lostpedia wiki to reveal the structure of topics being written about, the trends in attention devoted to topics, and relating the model inferences to events known to have occurred in the real world. Relevance of the models to progress in exploratory search and social information foraging is discussed.
Subjects: 4. Cognitive Modeling; 6. Computer-Human Interaction
Submitted: Jan 25, 2008