Terminological Reasoning with Knowledge and Beliefs

Hans-Jürgen Bürckert, Andrea Gräber, Armin Laux

In: Armin Laux , Heinrich Wansing (Hrsg.). Knowledge and Beliefs in Philosophy and Artificial Intelligence. Seiten 29-64 Akademie-Verlag 1995.


Terminological knowledge representation languages in the style of Kl-ONE are used to represent the taxonomic and conceptual knowledge of a problem domain in structured and well-formed way. However, representing knowledge of an application domain with such a formalism amounts to describing the relevant part of the "world" by listing the facts that hold in this part of the world. Subjective information of agents - which is essential for the representation of multi-agent environments - can only be expressed in a very limited way. For the representation of subjective information, modal logics with possible worlds semantics provides a formally well-founded and well-investigated basis.
In this article we show how modal logics can be employed to integrate the epistemic operators knowledge and belief into terminological knowledge representation languages. We introduce syntax and semantics of the extended language, and show that satisfiability of finite sets of formulas is decidable. In particular, it is taken into consideration that knowledge and belief of agents may influence each other, and this interaction is described by means of so-called interaction axioms.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence