|
Expert System is an artificial intelligence technology based on knowledge, which imitates expert decision based on reasoning and judgment according to special knowledge and experience provided by expert in a certain domain. As a‘shallow knowledge’system, the capability of expert system is badly rely on the integrality of the knowledge-database and reasoning technique. As a computer expression of expert experience and knowledge, knowledge representation affects the knowledge acquisition, the integrality of the knowledge-database and reasoning technique directly. While traditional knowledge representation, such as production representation, exists many limitations. In the traditional production system, there are many challenges in the methods of knowledge acquisition, resource sharing, knowledge expansion, and function integration. So it is worth further work in this field.Firstly, the application, status quo of expert system are briefly introduced in this paper, and the current deficiencies of production system are analyzed in the aspect of application; Then, a semantic element model based on production support the acquisition and expression are derived, and meta expression, semantic, formalization define and formalization describe are given. Secondly, in order to make up the lack of knowledge acquisition, resource sharing, knowledge expansion, we proposed a new expert system architecture with the scheme of model driven architecture, and particularly introduced implement graphical modeling tool and code generation engine; Meanwhile, in order to realize the forward and backward reasoning on the FuzzyCLIPS engine, a backward knowledge chain is presented and a mapping from knowledge based on semantic element model to backward chain is designed; Finally, the paper introduces the realization and application of‘Health Evaluating and Disease Analyze System’, which provided by China Aerospace Science & Industry Second Academy. The well running of the system proves the feasibility of the method.
|