COOP: A SELF-ASSESSMENT BASED APPROACH TO COOPERATING EXPERT SYSTEMS
Shashi Shekhar, C. V. Ramamoorthy · International Journal of Artificial Intelligence Tools · 1992
Conventional Expert System Shells do not help in developing AI programs for large applications like automated factories, which require multi-disciplinary knowledge and which are geographically distributed. To support these applications, a shell must provide tools for a knowledge-based system to (i) reason about the need for cooperation, (ii) understand global knowledge to locate relevant expert systems and (iii) select appropriate cooperation plans. Contemporary approaches like Blackboard [1], Contract-net [2] and Distributed problem solving [3] help in exploring alternative cooperation plans without any reasoning about the need for cooperation and understanding of global knowledge. Coop [4] support cooperation models to characterize three essential decisions in the cooperation process. It provides a computational method to decide if an expert system needs to consult with other expert systems. We provide techniques select appropriate cooperation plans.