Asymptotically Idempotent Aggregation Operators for Trust Management in Multi-agent Systems

The study of trust management in multi-agent system, especially distributed, has grown over the last years. Trust is a complex subject that has no general consensus in literature, but has emerged the importance of reasoning about it computationally. Reputation systems takes into consideration the history of an entity’s actions/behavior in order to compute trust, collecting and aggregating ratings from members in a community. In this scenario the aggregation problem becomes fundamental, in particular depending on the environment. In this paper we describe a technique based on a class of asymptotically idempotent aggregation operators, suitable particulary for distributed anonymous environments.