ABSTRACT
The measurement of expert systems development efforts and systems challenges existing technologies for software assessment. We view an expert system as an artificial intelligence-based program that performs a task that a human expert can accomplish. The challenges for assessment stem from the knowledge engineering approach to software development and the nature of the expert system applications typically developed. Standard approaches to verification, validation, and quality assurance do not adequately address these challenges (Geissman & Schultz, 1988; Swigger, chap. 6).
