Active set model for agent consensus and integration
Germano Resconi · 2012
In knowledge processing we must solve the problem to eliminate inconsistency, find consensus and make integration. Agents in their vote process are in a conflict state because they are not in agreement on the definition of the same object. Agents have not always consensus how to integrate their knowledge to build one new knowledge. Agents and the universe of objects must be compared in a way to discover conflicts and make process to generate consensus and integration. We suggest that sets of agents are not ordinary sets but are active sets with their proper logic. With the definition of the conflict profile we can define special active sets that solve the conflict problem. We see that the minimum distance is not the only criterion to solve the conflict. We must check also the consensus of all the agents inside their active set. Integration as knowledge agreement is defined in this paper but not solved.