Reasoning about Weighted Semantic User Profiles through Collective Confidence Analysis: A Fuzzy Evaluation
2010 (English)In: ADVANCES IN INTELLIGENT WEB MASTERING-2, PROCEEDINGS / [ed] Snasel V; Szczepaniak PS; Abraham A; Kacprzyk J, 2010, Vol. 67, 71-81 p.Conference paper (Refereed)
User profiles are vastly utilized to alleviate the increasing problem of so called information overload. Many important issues of Semantic Web like trust, privacy, matching and ranking have a certain degree of vagueness and involve truth degrees that one requires to present and reason about. In this ground, profiles tend to be useful and allow incorporation of these uncertain attributes in the form of weights into profiled materials. In order to interpret and reason about these uncertain values, we have constructed a fuzzy confidence model, through which these values could be collectively analyzed and interpreted as collective experience confidence of users. We analyze this model within a scenario, comprising weighted user profiles of a semantically enabled cultural heritage knowledge platform. Initial simulation results have shown the benefits of our mechanism for alleviating problem of sparse and empty profiles.
Place, publisher, year, edition, pages
2010. Vol. 67, 71-81 p.
, Advances in Intelligent and Soft Computing, ISSN 1867-5662 ; 67
Confidence, Fuzzy Inference, Semantic User Profiles, Personalization, Reasoning, Uncertainty Evaluation
Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-29226DOI: 10.1007/978-3-642-10687-3_7ISI: 000281727500007ScopusID: 2-s2.0-84886024779OAI: oai:DiVA.org:kth-29226DiVA: diva2:392960
6th Atlantic Web Intelligence Conference, Charles Univ, Fac Math & Phys, Prague, CZECH REPUBLIC, SEP, 2009
QC 201101282011-01-282011-01-272013-02-19Bibliographically approved