Trust Inference Path Search with Minimum Uncertainty for E-Commerce

Trust inference is a common method for trusted decision-making between two unfamiliar users in E-commerce environments. A challenging preceding work of trust inference is to find appropriate trust propagation paths. Existing trust network discovery approaches mainly apply the classical graph searching algorithms and cannot provide reliable trust inference paths. To solve this issue, we first analyze the trust discounting operators with structure balance theory and investigate the relationship between balanced transitive triads and edge uncertainties. Then, the Minimum Inference Uncertainty Search (MIUS) strategy is formally described and the trust inference path search algorithm is proposed on the basis of the bidirectional versions of Dijkstra´s algorithm. The comparative experiments of path search and sign prediction are conducted on the Epinions data set. The experimental results show that the proposed algorithm can find the trust inference path with higher efficiency and the path has better applicability for trust inference than the existing ones.