Representation and reasoning methodology for constraint and rule satisfaction problems

Abstract

Numerous rule-based expert systems are successfully applied to field problems, while more attention of Artificial Intelligence researchers is given to Constraint Satisfaction Problems (CSP). Recent research on CSP attempts to improve CSP representation and solving algorithms for real world problems. So it``s time to integrate rules with constraints in a unified framework. We propose a hybrid representation of rules and constraints along with so called hybrid reasoning for our target problem - Constraint and Rule Satisfaction Problems (CRSP). The hybrid reasoning has the following three characteristics : 1) supporting user interaction to reflect user intention about the problem situation and to resolve conflicts between objectives, 2(concurrent reasoning with multiple starting points to enhance the mearch efficiency, 3) integrated reasoning encompassing the backward chaining popularly used in the rule-based systems and constraint propagation methods developed for CSP. The CRSP framework is applied to the planning of expert system development. For expert system development planning, there has been yet no systematic, computerized tool to generate a feasible plan. ES development planning problems need both constraints and rules for representation, and the proposed hybrid reasoning can provide a feasible plan. We can show the applicability of CRSP framework with the illustrative tax advisory ES development planning. A prototype system UNIK-CRSP is developed to realize the proposed representation and reasoning. The CRSP framework seems applicable to many other planning and design problems, like configuration, system integration, and information system planning.