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.
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.
P. Cimiano, A. Hotho, G. Stumme, and J. Tane. Concept Lattices, volume 2961 of LNAI, page 189-207. Heidelberg, Second International Conference on Formal Concept Analysis, ICFCA 2004, Springer, (2004)
M. Rese, and U. Bretschneider. Proceedings Paderborner Frühjahrstagung: Supply Chain Management in der Automobil- und Zulieferindustrie, page 129-138. Paderborn, Germany, (2004)
A. Ulbrich, G. Mühl, T. Weis, and K. Geihs. Confederated International Conferences CoopIS, DOA, and ODBASE 2004, (2004)Lecture Notes in Computer Science (LNCS), Larnaca, Cyprus, October 2004. Springer..
K. Herrmann, K. Geihs, and G. Mühl. IFIP TC8 Working Conference on Mobile Information Systems (MOBIS 2004), (2004)Oslo, Norway, September 2004. Kluwer.
H. Krcmar, Y. Arnold, M. Daum, and J. Leimeister. ACM Group 2003 Conference International Conference on Supporting Group Work, Sanibel Island, Florida, USA, (2003)EXT29.
M. Daum, A. Klein, J. Leimeister, and H. Krcmar. Arbeitspapier Nr. 109 des Lehrstuhls für Wirtschaftsinformatik der Universität Hohenheim, 109, Universität Hohenheim, Hohenheim, Germany, (2001)