Graded Belief Revision for Jason: A Rule-Based Approach

Jason is a Java-based platform for the development of multi-agent systems, which is a particular implementation of AgentSpeak. While some theoretical proposals have been put forward to add to Jason both belief revision and the preference order on the agent's beliefs, the reasoning on the practical way to integrate such proposals as well as their implementation have not been considered. This paper aims to fill those gaps, by adding the concept of graded beliefs and making use of Jason customisation features to implement reasoning and belief revision capabilities. The resulting approach allows agents to reason about the belief's degree of certainty, track dependency between them, and revise the belief set accordingly. A running example will illustrate the presented work and highlight its added value.