Anon Plangprasopchok, Kristina Lerman
Tagging in social media system has demonstrated to be a convenient way for users to annotate objects of interest.One reason behind its success obviously because tags can be chosen by users arbitrarily without any topic and specificity constraints. Although tags are free-from keywords, there are some evidences suggesting that, for a particular object type, users tend to use ``similar'' tag sets. In addition, such tags are in different levels of specificity. This might suggest that there are some hierarchical concepts behind users' tagging processes. In this paper, we outline a problem in extracting hierarchical concepts from social annotation data and propose a possible solution -- a probabilistic generative model that describes tagging processes.
Subjects: 12. Machine Learning and Discovery; 3.4 Probabilistic Reasoning
Submitted: Jan 24, 2008