A RECOMMENDER MODEL FOR SOCIAL BOOKMARKING SITES

Social bookmarking and other Web sites allow users submitting their resources and labeling them with arbitrary keywords, called tags, to create folksonomies. These sites usually provide their users tag recommendations in order to help them to find relevant information and resources. However, only very basic techniques are applied for generating recommendations. In this paper, we present a recommender system for a social bookmarking site to generate resource recommendations rather than tag recommendations. Our system is based on two ideas: similar users are interested in similar resources and similar resources have similar tags. Our system generates recommendations by automatically taking into account what resources a user tags and the co-occurrence of tags. Our method is tested on large-scale real life datasets. The experimental results show that our method achieves a good recommendation performance.