A Principled Approach to Knowledge-Based System Development
- Guus Schreiber
- Bob Wielinga
- Joost Breuker
KADS is a structured methodology for the development of knowledge based systems which has been adopted throughout the world by academic and industrial professionals alike. KADS approaches development as a modeling activity. Two key characteristics of KADS are the use of multiple models to cope with the complexity of knowledge engineering and the use of knowledge-level descriptions as an immediate model between system design and expertise data. The result is that KADS enables effective KBS construction by building a computational model of desired behavior for a particular problem domain.KADS contains three section: the Theoretical Basis of KADS, Languages and Tools, and Applications. Together they form a comprehensive sourcebook of the how and why of the KADS methodology. KADS will be required reading for all academic and industrial professionals concerned with building knowledge-based systems. It will also be a valuable source for students of knowledge acquisition and KBS.
AUDIENCE KBS and AI professionals, software engineers, academic research level AI and systems engineering.
Hardbound, 457 Pages
Published: March 1993
Imprint: Academic Press
- Introduction and Overview. Theoretical Basis. Modeling Expertise. Modleing System: User Cooperation in KADS. A KADS Domain Description Language. Model Construction. Operationalizing Models of Expertise. KADS and Conventional Software Engineering. Languages and Tools. A Formal Language for KADS Models of Expertise. MODEL-K: Making KADS Run. Using OMOS to Represent KADS Conceptual Models. Shelley-Computer Aided Knowledge Engineering. Applications. The Paint Advisor: A Communicative Expert System. Managing and Maintaining an Operational KADS System. Rational Reconstruction of Diagnostic Expertise. Qualitative Prediction of Behavior. A KADS Conceptual Model of the ONOCIN Task. Applying KADS to the Sisyphus Domain. Using KADS to Analyze Problem-Solving Methods. Bibliography. Name Index. Subject Index.