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.
T. Hanika, и T. Hille. Conceptual Knowledge Structures - First International Joint Conference, CONCEPTS 2024, Cádiz, Spain, September 9-13, 2024, Proceedings, том 14914 из Lecture Notes in Computer Science, стр. 97--112. Springer, (2024)
M. Stubbemann, и G. Stumme. Machine Learning and Knowledge Discovery in Databases: Research Track - European Conference, ECML PKDD 2023, Turin, Italy, September 18-22, 2023, Proceedings, Part III, том 14171 из Lecture Notes in Computer Science, стр. 177--192. Springer, (2023)
M. Stubbemann, T. Hanika, и G. Stumme. Advances in Intelligent Data Analysis XVIII - 18th International Symposium on Intelligent Data Analysis, IDA 2020, Konstanz, Germany, April 27-29, 2020, Proceedings, том 12080 из Lecture Notes in Computer Science, стр. 496--508. Springer, (2020)
M. Stubbemann, и G. Stumme. Machine Learning and Knowledge Discovery in Databases: Research Track - European Conference, ECML PKDD 2023, Turin, Italy, September 18-22, 2023, Proceedings, Part III, том 14171 из Lecture Notes in Computer Science, стр. 177--192. Springer, (2023)
M. Stubbemann, и G. Stumme. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2023, том 14171 из Lecture Notes in Computer Science, стр. 177--192. Springer, (2023)
B. Ganter, T. Hanika, и J. Hirth. Formal Concept Analysis - 17th International Conference, ICFCA 2023, Kassel, Germany, July 17-21, 2023, Proceedings, том 13934 из Lecture Notes in Computer Science, стр. 64--77. Springer, (2023)
B. Ganter, T. Hanika, и J. Hirth. Formal Concept Analysis - 17th International Conference, ICFCA 2023, Kassel, Germany, July 17-21, 2023, Proceedings, том 13934 из Lecture Notes in Computer Science, стр. 64--77. Springer, (2023)
J. Hirth, V. Horn, G. Stumme, и T. Hanika. Graph-Based Representation and Reasoning - 28th International Conference on Conceptual Structures, ICCS 2023, Berlin, Germany, September 11-13, 2023, Proceedings, 14133, стр. 138--152. (2023)
T. Hanika, и J. Hirth. Graph-Based Representation and Reasoning - 26th International Conference on Conceptual Structures, ICCS 2021, Virtual Event, September 20-22, 2021, Proceedings, том 12879 из Lecture Notes in Computer Science, стр. 105--118. Springer, (2021)
T. Hanika, и J. Hirth. Graph-Based Representation and Reasoning - 26th International Conference on Conceptual Structures, ICCS 2021, Virtual Event, September 20-22, 2021, Proceedings, том 12879 из Lecture Notes in Computer Science, стр. 105--118. Springer, (2021)
G. Klumbyte, P. Lücking, и C. Draude. Proceedings of the 11th Nordic Conference on Human-Computer Interaction: Shaping Experiences, Shaping Society, New York, NY, USA, Association for Computing Machinery, (2020)
G. Klumbyte, C. Draude, и A. Taylor. CHItaly 2021: 14th Biannual Conference of the Italian SIGCHI Chapter, New York, NY, USA, Association for Computing Machinery, (2021)
T. Hanika, и J. Hirth. Formal Concept Analysis - 16th International Conference, ICFCA 2021, Strasbourg, France, June 29 - July 2, 2021, Proceedings, том 12733 из Lecture Notes in Computer Science, стр. 261--269. Springer, (2021)
L. Stubbemann, D. Dürrschnabel, и R. Refflinghaus. (2021)cite arxiv:2103.10451Comment: 16 pages, 6 figures, 1 table, Accepted to: ETRA2021, ACM Symposium on Eye Tracking Research and Applications.
D. Dürrschnabel, и G. Stumme. (2021)cite arxiv:2102.02684Comment: 16 pages, 6 figures, 4 algorithms, for source code refer to https://github.com/domduerr/redraw.
T. Hanika, F. Schneider, и G. Stumme. Accepted for publication in: Tohoku Mathematical Journal, (2020)cite arxiv:1801.07985Comment: v2: completely rewritten 28 pages, 3 figures, 2 tables.