S-BITS : Social-Bookmarking Induced Topic Search

With the recent flood of information, trust in that information is gaining a lot of attention in determining how information is used. Social bookmarking is a new information sharing service that allows individuals to bookmark and annotate web pages of interest or those that impress them. It is attracting attention and growing in popularity. In social bookmark services, users´ bookmarks and annotations given by tags are informative indicators of user interest in web pages. This paper proposes a method to evaluate the trust and significance of web pages based on social bookmarks. Extending the HITS approach, we regard web pages as Authority and users as Hubs and evaluate trust and significance values of web pages. We show usefulness of the proposed approach through experiments.