Abstract:More and more users joined in social network. The precise social trust value is critical for application system such as recommendation system. To a user, the egocentric network is formed by the user, his friends and social relationships between him and other users. We proposed an algorithm for inferring dynamic trust based on trust chains and interactions. Indirect trust values are calculated depending on direct trust values and trust chains in the egocentric network. As the social network evolves, the dynamic trust values can be resulted from the interactions between a user and his friends and trust reference. By statistical analysis of trust value in social network, this algorithm improved the accuracy of trust transitivity and trust value computing compared with a classical trust algorithm.

A Dynamic Trust Conference Algorithm for Social Network