Knowledge-based systems: A qualitative formalism

This paper examines knowledge-based systems (KBS's) and considers how they may be applied in the social sciences. KBS's are programs which use artificial intelligence techniques to solve complex problems. Characteristics of KBS's are illustrated with an example knowledge-based program (N-ACT) written in PROLOG which uses Goffman's dramaturgical model to analyze social interaction. The paper argues that KBS's offer a qualitative formalism capable of being applied to concepts and phenomena which have been beyond the scope of traditional mathematical models. KBS's can create explicit, reproducible, logically powerful knowledge systems while avoiding the Procrustean bed of mathematical formalism.This research took place in part while the author was on sabbatical leave from the University of Missouri and a post-doctorate fellow in medical computing and informatics on NIH Grant LM 07006 from the National Library of Medicine. The author would like to thank Peter H. Hall and two anonymous reviewers for their helpful comments on earlier drafts.