Deep Interoperation betweenPrevious: Introduction
Developed expert systems are much different in their implementation details, such as knowledge representation languages and how to run inference engines. If they would exchange information at such implementation details, each expert system would have no way to find information applicable to its better performance from other expert systems. On the contrary, if they exchange information at a too general level, the available information would be missing for each expert system to improve its performance. So, in order for distributed diagnostic expert systems to exchange useful information, they should be modeled and specified at a conceptually acceptable level of detail. The following is a method to identify the proper kinds and granularity of information for distributed diagnostic expert systems to be modeled, from the point of knowledge level analysis.