JoWUA

Abstract

Social networks express the information flows of individuals or groups using actors and relationships. Along with the growth of WWW, interest in large-scaled social networks has grown bigger. However, social networks being applied in diverse areas are facing with sparsity problems resulting from a shortage of actual relations with structural properties. Therefore, in this paper, sparsity problems are solved by creating contents based virtual relations applied with the clustering approach. In addition, a reliable social trust model is presented through combinations of actual relations and virtual relations. In experiments, the performance of the proposed approach will be evaluated using connectivity between nodes and F-measures.