Exploiting the social capital of folksonomies for web page classification

Abstract

Collaborative tagging systems (CTSs), also known as folksonomies, have grown in popularity on the Web and social tagging has become an important feature of many Web 2.0 services. It has been argued that the power of tagging lies in the ability for people to freely determine the appropriate tags for resources without having to rely on a predefined lexicon or hierarchy. The freeform nature of tagging causes a number of problems in this social classification scheme, such as synonymy and morphological variety. However, social tagging can be a valuable source of information to help in the organization of Web resources. In this paper we present an empirical analysis carried out to determine the importance of social tagging in Web page classification. Experimental results showed that tag-based classification outperformed classifiers based on full-text of documents. © 2010 IFIP International Federation for Information Processing.