Subsumption Algorithms for Some Attributive Concept Description Languages

Abstract

This paper investigates subsumption algorithms for logic-based knowledge representation languages of the KL-ONE family. We amalgamate the attributive concept description language ALC, that contains value restrictions, intersections, unions and complements with number restrictions, role hierarchies (to model the KL-ONE’s roleset differentiation), and Feature Logic, respectively. We show that deciding consistency and subsumption of ALC extended with number restrictions and ALC extended with role hierarchies is PSPACE-complete. Furthermore, for all these languages we give subsumption algorithms.