On integrating expert system programs into cooperative systems

Abstract

The author points out an opportunity for using first-generation expert system programs to construct systems that can solve large-scale problems in a cooperative manner. He describes the requirements for expert system programs to support integration into such systems and proposes the application of the communication-free cooperating agents theory (M.R. Genesereth et al., 1986) as a model for these systems. A minimal modification approach that uses an integration agent is discussed, and the feasibility of the proposal is assessed. An illustrative example from the design domain is included.<>