Transparent Ontological Integration of Multi-Agent Systems

One of the main problems that are faced in the integration of multi-agent or holonic systems is the ontological heterogeneity. Two agents are able to communicate only when they understand ontologies used in exchanged messages. This problem can be solved by translating messages between different ontologies during the communication. Still, there is the question of how to integrate such a translation service into the whole system. There are several proposals, such as requiring each agent to be able to do the translation, or using specialized ontology service. However, none of these approaches is feasible for the integration of simple agents that are hard to be considerably changed and adapted to process multiple ontologies. In this paper, we discuss architectural possibilities and propose a way of transparent translation during agent communication, where agents do not even need to be aware of ontology processing.