Ontological Feedback in Multiagent Systems1

When agents transfer information, they need a conceptualisation of the domain of interest and a shared vocabulary to communicate facts with respect to this domain. The conceptualisation can be expressed in a so-called ontology, which is often defined in a formal language, such as a programming language or a formal logic. An ontology abstracts the essence of the domain of interest and helps to catalogue and distinguish various types of objects in the domain, their properties and relationships (see, e.g. [3, 4]). In successful communication and in collaborative performance of tasks, agree-ment between different agents with respect to the ontology is crucial or, at least, the agents should be aware of existing discrepancies. In case of discrepancies, var-ious troublesome situations may arise. One of the agents, for instance, by mistake may assume that a particular concept is shared, while the other has no knowledge about it. Or worse, both dialogue partners have different conceptualisations, while the relevant discrepancy remains unnoticed. Ontological discrepancies may cause serious communication flaws and the generation of adequate feedback in order t