Validation and Verification of Knowledge-Based Systems
Papers from the AAAI Workshop
Alun D. Preece, Program Chair
For as long as knowledge-based systems (KBS) have been built, establishing their reliability has been known to be a difficult problem. However, in the last few years realistic techniques have begun to appear which address this problem. The specific objectives of this year's workshop are to continue the evaluation of practical approaches and issues in validation and verification (ideally supported by empirical data) and to encourage the growing maturity of the field by raising awareness of the relationships between V&V field and other aspects of AI and software engineering. Specific focus of this technical report will includea nalyses of the effectiveness of validation and verification techniques used in practice; relationships between knowledge-based systems and conventional V&V techniques; the role and use of meta-knowledge in knowledge-based system validation and verification processes; validation and verification of embedded KB; and an investigation of potential regulatory activities for and with validation and verification.