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.
F. Mitzlaff, D. Benz, G. Stumme, and A. Hotho. Proceedings of the 21st ACM conference on Hypertext and hypermedia, page 265--270. New York, NY, USA, ACM, (2010)
M. Atzmueller, and F. Lemmerich. Proc. ECML/PKDD 2012: European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Heidelberg, Germany, Springer Verlag, (2012)
M. Atzmueller, and S. Beer. Proc. 55th IWK, International Workshop on Design, Evaluation and Refinement of Intelligent Systems (DERIS), University of Ilmenau, (2010)
C. Scholz, M. Atzmueller, and G. Stumme. Proc. 21st Intl. Symposium on Methodologies for Intelligent Systems, Heidelberg, Germany, Springer Verlag, (2014)
M. Atzmueller, and F. Mitzlaff. Workshop on Mining Patterns and Subgroups, Lorentz Center, Leiden, The Netherlands. Awarded with the Best Discovery Award, (2010)
M. Atzmueller, and F. Mitzlaff. Lorentz Workshop on Mining Patterns and Subgroups, Lorentz Center, Leiden, The Netherlands. Awarded with the Best Discovery Award, (2010)
M. Atzmueller, S. Bobek, M. Kibanov, and G. Nalepa. Proceedings of the 2014 International Smart University Workshop (SmartU 2014), London, UK, London, UK, (2014)
M. Atzmueller, and K. Hilgenberg. Proc. 4th International Workshop on Modeling Social Media (MSM 2013), Hypertext 2013, New York, NY, USA, ACM Press, (2013)
S. Doerfel. Concept Lattices and Their Applications 2010 - Proceedings of the 7th International Conference on Concept Lattices and Their Applications Sevilla, Spain, October 19-21, 2010., 672, page 283-293. CEUR-WS, (November 2010)