Balance Theory-Based Model for Discovering Trust Network

Relations among users on social network often reflect a mixture of positive (trust) and negative (distrust) interactions. It is necessary to extract the trust relationships among people and discover a trust network from the complex social network. The trust network can be used to calculate the trust confidence and can infer the trust values among the nodes in it. In this paper, we propose a method to estimate a trust network and it is the first time that the balance theory is applied to solve the problem of discovering trust network. The method makes use of balance theory to estimate the credibility of the relationships among the nodes. First, we analyzed the core content of the balance theory and the similarities with trust network. Then, we proposed the rules of building the trust network, and summarized the process of the implementation. The experiment runs on Epinions data set which consists of more than 130,000 nodes. The experiment results demonstrate that our algorithm performs well in building trust network.