One-class classification for finding interesting resources in social bookmarking systems

DANIELA GODOY

Paris

Workshop; 3rd International Workshop on Resource Discovery (RED 2010) at the 12th International Conference on Information Integration and Web-based Applications & Services (iiWAS2010); 2010

In this work two one-class classification approaches were empirically evaluated and compared for the task of identifying interesting resource in social tagging systems. One-class SVM and Rocchio classifiers were used to learn the user interests starting from different sources, such as the full-text of resources and their social tags.