K-Tree: An Efficient Structure for Verification and Inference in Rule-Based Systems

A K-Tree based approach for verifying completeness and consistency in rule-based systems is discussed. A structure called Parameter Dependency Network (PDN) is used to perform the global level verification. The K-Tree approach has the potential to carry out verification, in an efficient manner than alternative approaches, and it can be generally implemented in any rule-based expert system shell. A K-Tree Based Inference Algorithm (KTBIA) for rule-based systems is also presented. KTBIA makes use of the K-Tree and PDN structures generated during the verification process.