Verification and Validation of Embedded Knowledge-Based Software Systems

摘要

Our overall goal in this research effort has been to reduce the time and cost of constructing embedded knowledge based systems that must handle uncertainty in information in a rigorous manner. Our fundamental approach actively assists subject-matter experts in organizing their knowledge inclusive of uncertainty to build such embedded systems in a consistent and correct as well as effective fashion. We pursued this by carefully examining the nature of uncertainty and information semantics and developing intelligent tools for verification and validation that provides assistance to the subject-matter expert in constructing their knowledge based systems. We have developed a prototype environment for constructing Bayesian Knowledge Bases called PESKI.