CISC 881: Knowledge-Based Systems

Catalog Description:
Architecture of knowledge-based systems, development and application of knowledge bases in systems for expert performance in specialized domains. Knowledge sources, rule-based inference, uncertainty of facts and explanation of reasoning.

Current Texts:
Building Expert Systems
Frederick Hayes-Roth, Donald A. Waterman and Douglas B. Lenat (editors)
Addison-Wesley Publishing Company, 1983

Goals:
The student will learn how concepts of artificial intelligence are applied to perform practical tasks with the skill of an expert, how expert systems are organized, and how expert knowledge is put into them.

Content:

• Concepts of artificial intelligence applied to practical tasks requirin g large amounts of knowledge of the task domain
• Possible architectures for knowledge-based systems are examined, including systems based on backward-chaining, production rules, frames and constraint satisfaction.
• Techniques of "knowledge engineering'' will be emphasized: multiple knowledge sources, rule-based inference, focus of attention, uncertainty of facts, hypothetical reasoning, explanation of responses, automatic discovery of facts, metaknowledge, reasoning about structure and function, and causal models.
• System development is also covered, from application selection and system design, to knowledge acquisition, testing and evaluation, and commercialization
• Discussion of tools and languages recommended for knowledge-based system development
• Participation in class includes building a small knowledge-based system.

Required Background: CISC 681 (Artificial Intelligence) or consent of the instructor.

Helpful Background: Knowledge of LISP