Social bookmarking systems and their emergent information structures, known as folksonomies, are increasingly important data sources for Semantic Web applications. A key question for harvesting semantics from these systems is how to extend and adapt traditional notions of similarity to folksonomies, and which measures are best suited for applications such as navigation support, semantic search, and ontology learning. Here we build an evaluation framework to compare various general folksonomy-based similarity measures derived from established information-theoretic, statistical, and practical measures. Our framework deals generally and symmetrically with users, tags, and resources. For evaluation purposes we focus on similarity among tags and resources, considering different ways to aggregate annotations across users. After comparing how tag similarity measures predict user-created tag relations, we provide an external grounding by user-validated semantic proxies based on WordNet and the Open Directory. We also investigate the issue of scalability. We find that mutual information with distributional micro-aggregation across users yields the highest accuracy, but is not scalable; per-user projection with collaborative aggregation provides the best scalable approach via incremental computations. The results are consistent across resource and tag similarity.
S. Doerfel, D. Zoller, P. Singer, T. Niebler, A. Hotho, and M. Strohmaier. Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion, page 251-252. Republic and Canton of Geneva, Switzerland, International World Wide Web Conferences Steering Committee, (2014)
F. Mitzlaff, S. Doerfel, A. Hotho, R. Jäschke, and J. Mueller. 15th Discovery Challenge of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2013, Prague, Czech Republic - October 27, 2013. Proceedings, 1120, page 7-24. Aachen, Germany, CEUR-WS, (2014)
S. Doerfel, R. Jäschke, A. Hotho, and G. Stumme. Proceedings of the 4th ACM RecSys workshop on Recommender systems and the social web, page 9--16. New York, NY, USA, ACM, (2012)
S. Doerfel, R. Jäschke, and G. Stumme. Formal Concept Analysis - 10th International Conference, ICFCA 2012, Leuven, Belgium, May 7-10, 2012. Proceedings, volume 7278 of Lecture Notes in Computer Science, page 77-95. Berlin/Heidelberg, Springer, (2012)