Improving folksonomies using formal knowledge: a case study on search

Full text available as:

PDF (Version of Record) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader Download (439Kb) | Request Copy from OU Author

Abstract

Search in folksonomies is impeded by lack of machine understandable descriptions for the meaning of tags and their relations. One approach to addressing this problem is the use of formal knowledge resources (KS) to assign meaning to the tags, most notably WordNet and (online) ontologies. However, there is no insight of how the different characteristics of such KS can contribute to improving search in folksonomies. In this work we compare the two KS in the context of folksonomy search, first by evaluating the enriched structures and then by performing a user study on searching the folksonomy content through these structures. We also compare them to cluster-based folksonomy search. We show that the diversity of ontologies leads to more satisfactory results compared to WordNet although the latter provides richer structures. We also conclude that the idiosyncrasies of folksonomies can not be addressed by only using formal KS.