Users in social network use social tagging to annotate the free text to the resource in terms of metadata, which describes users' characteristics or interests. However, the majority of current recommender systems use the users' tagging behaviors to recommend the resource by focusing on what users tag or direct interests only. This research presents an approach to capturing user direct and indirect interests using Virtual Tag Cloud (VTC) for the tag-based recommender systems. The manipulation of VTC and the resource recommendation model are described in this work. The assessment of the effectiveness of the proposed method is measured by F-Measure.
