Research of a Multi-dimension Measurement Trusting Model Based on Markov Chain

There are uncertain, dynamic and subjective features about trust relation in realistic network society, how to accurately depict and compute the trust values is a challenging issue in trust computing. Integrating with practical e-commerce application, this paper introduces a novel multi-dimension trusting model based on Markov chain mechanism consisting of evaluating, calculating, filtrating, and publication center. The multi-dimension evaluation metrics include the functional and nonfunctional properties and the corresponding weights. We adopt the continuous evaluation values to measure the trustworthiness of services. The Markov´s chain-based trust measurement is given to dynamically calculate the trust value and discover the malicious behaviors. In order to conveniently manage and avoid the dead-lock, some constraint rules are proposed.