Dart: a Distributed Analysis of Reputation and Trust Framework

Artificial societies distributed systems of autonomous agents are becoming increasingly important in open distributed environments, especially in e-commerce. Agents require trust and reputation concepts in order to identify communities of agents with which to interact reliably. We have noted in real environments that adversaries tend to focus on exploitation of the trust and reputation model. These vulnerabilities reinforce the need for new evaluation criteria for trust and reputation models called exploitation resistance which reflects the ability of a trust model to be unaffected by agents who try to manipulate the trust model. To examine whether a given trust and reputation model is exploitation resistant, the researchers require a flexible, easy-to-use, and general framework. This framework should provide the facility to specify heterogeneous agents with different trust models