Common Sense Reasoning - Towards Mature Qualitative Reasoning Engines

Samenvatting

Qualitative Reasoning captures human common sense understanding of the behavior of real world physical systems. Qualitative models use a rich explicit conceptual vocabulary and therefore a qualitative simulation provides an explanation as well as a description of the system's behavior. These properties make Qualitative Reasoning suitable for use in several application fields like education and ecology. The design of an elaborate, detailed Qualitative Reasoning engine is a complex task. Existing simulators lack reasoning capacities which are required by present day applications, such are extensive use of inequality reasoning and efficient computation of state transitions. This thesis describes an effort to develop fundamental solutions for several problems in such reasoning engines, particularly focussing on the GARP reasoning engine.