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.
R. Kniewel, M. Radziwill, C. Evers, and L. Schmidt. Useware 2014: Interaktionen mit großen Datenmengen und unscharfen Analyseergebnissen (Heilbronn 2014), page 121–131. Düsseldorf, VDI-Verlag, (2014)
M. Heußner, L. Ackermann, and L. Schmidt. Wohnen - Pflege - Teilhabe: Besser leben durch Technik : 7. Deutscher AAL-Kongress (Berlin 2014), page 1-6 (a40). Berlin, VDE-Verlag, (2014)
K. Behrenbruch, M. Söllner, J. Leimeister, and L. Schmidt. Human-Computer Interaction: 14th IFIP TC 13 International Conference (Kapstadt 2013), Part IV, page 306–313. Springer, Berlin, (2013)
K. Behrenbruch, and L. Schmidt. Grundlagen und Anwendungen der Mensch-Maschine-Interaktion: 10. Berliner Werkstatt Mensch-Maschine-Systeme (Berlin 2013), page 117–125. Berlin, Universitätsverlag der TU Berlin, (2013)
L. Ackermann, M. Heußner, and L. Schmidt. 11. E-Learning Fachtagung Informatik der Gesellschaft für Informatik e. V. (Bremen 2013), page 257–262. Bonn, Köllen, (2013)
D. Ziegner, R. Herrmann, J. Hegenberg, and L. Schmidt. Gestaltung der Arbeitswelt der Zukunft: 60. Kongress der Gesellschaft für Arbeitswissenschaft (München 2014), page 115–117. Dortmund, GfA-Press, (2014)
L. Schmidt, J. Hegenberg, B. Borys, and R. Kniewel. Useware 2014: Interaktionen mit großen Datenmengen und unscharfen Analyseergebnissen (Heilbronn 2014), page 11–31. Düsseldorf, VDI-Verlag, (2014)
M. Radziwill, and L. Schmidt. Gestaltung der Arbeitswelt der Zukunft: 60. Kongress der Gesellschaft für Arbeitswissenschaft (München 2014), page 91–93. Dortmund, GfA-Press, (2014)
B. Löffler, and L. Schmidt. Forum Arbeitsphysiologie: 18. Symposium Arbeitsmedizin und Arbeitswissenschaft für Nachwuchswissenschaftler (Magdeburg 2014), page 25. Otto-von-Guericke-Universität Magdeburg, (2014)