An Approach for Detecting and Resolving Indirect Normative Conflicts in Multi-agent Systems.

In Multi-agent systems (MAS), norms can be adopt as a strategy to regulate and guide the behavior of software agents and avoid that unexpected actions occur in the system. In order to guarantee that the MAS runs properly, the set of norms must be free of conflict. Two norms are in conflict when the agent automatically violates a norm when adopts the other one. In this paper, we present an approach for detecting and resolving indirect conflicts among norms that regulate a multi-agent system. Indirect conflicts are those who arise among norms whose elements being regulated are not the same but are related. Our approach uses a lexical database and a domain ontology to detect indirect conflicts and the conflicts detected are resolved by manipulating the activation/deactivation conditions of the conflicting norms taking into account the relationships identified among the elements of the conflicting norms.