Semantic Enrichment of Folksonomies

Abstract

This thesis is set in the research effort to bridge Social Web (also called Web 2.0) with Semantic Web. In particular, we looked for ways of bridging Social tagging-based systems with structured representations such as thesauri or ontologies (in the informatics sense). Social tagging platforms allow their users to associate freely chosen signs to their favorite resources. These platforms have recently become very popular as a means to classify large sets of resources shared among on-line communities over the social Web. However, the folksonomies resulting from the use of these systems revealed limitations: tags are ambiguous and their spelling may vary, and folksonomies are difficult to exploit in order to retrieve or exchange information. The goal of this thesis is to overcome these limitations and to support the use of folksonomies with formal languages and ontologies from the Semantic Web, while proposing an approach to take the benefits of social dynamics found on the Web 2.0 for the elaboration of thesauri or ontologies. This thesis present our multi-points of view approach to the semantic enrichment of folksonomies. We propose a socio-technical system, grounded on a usage analysis, and combining automatic processing of tags and users contributions through user-friendly interfaces. Automatic processing of tags allows bootstrapping the process by using a combination of a custom method analyzing tags labels and adapted methods analyzing the structure of folksonomies. The contributions of users are described thanks to our model SRTag (Semantically Related Tag) that allows supporting diverging points of view, and captured thanks to our user friendly interface allowing the users to structure tags while searching the folksonomy. Conflicts arising between individual points of view are then detected and temporarily solved by an automatic agent, whose outcome is then exploited to help a referent user maintain a global and coherent structuring of the folksonomy. Each individual point of view can then be enriched with the others contributions, with the global point of view serving as a reference to guaranty a local coherence for all users. The result of our method allows enhancing the navigation within tag-based knowledge systems, but can also serve as a base for building thesauri or ontologies fed by a truly bottom up process, providing therefore a solution to the bottleneck effect of knowledge acquisition.