ALIGNING ONTOLOGIES IN MULTIPLE AGENTS' SOCIETY

摘要

A framework for aligning ontologies within the community of agents using different semantic similarity measures is proposed. The approach presented postulates the use of subjective logic for estimating concepts' correspondence and allows relating concepts from many ontologies. By means of subjective logic's consensus and recommendation operators the agents may exchange information about similarity of concepts and have possibility of relating concepts from non-aligned ontologies. Different strategies of achieving semantic interoperability within the agents' society are also discussed.