A Rough Set Approach to Agent Trust Management

Trust management in Multi Agent Systems plays a key role in today´s growing need for such systems. Methods of trust management try to approximate a set of best agents, in terms of some pre-defined measures. The theory of Rough Sets deals with the approximation of classifying objects in an environment for which we do not have a certain definition. In this research, we propose a new trust management approach based on the concepts of the theory of rough sets. We apply this approach on an Information Retrieval (IR) MAS. This application includes defining the attribute and their domain of values, discretizing the attribute values, collecting these values, and employing a rough set tool for analyzing them. The results of this application are presented.