Research And Implementation, Guizhou Province, Web-based Citrus Agricultural Expert System

Currently, the expert system based on Web technology is one of the very important projects which can be applied. The core of the system implementation is Knowledge Acquisition method, knowledge expression method, Knowledge Base Construction and reasoning technology. Thought the part of a National science and technology research project which is named Guizhou Citrus agricultural expert system (No. 2005BA901A04-4) based on Web. In this paper, we mainly research on the following aspects, such as: Knowledge Acquisition method based on BP artificial neutral network Algorithm, Knowledge Representation method based on weighted fuzzy production rules, Construction of XML database and the designation of fuzzy reasoning engine based on middleware technology, etc.As we all known, the reasoning engine is designed for Specific expert system and provide services only for this system in traditional expert system. When we want to design a new expert system, we must design a corresponding reasoning engine once again. Furthermore, as a close integration part of the expert system, the expansibility and reusability of reasoning engine is very badly. When the Knowledge Representation method be changed or improved, the system must be re-designed and re-realized. In order to solve this bottleneck problem of reasoning engine in expert system, a reasoning engine based on EJB (Enterprise Java Beans) technology is proposed in this paper. This object component of the intelligent reasoning engine has the following Benefits, such as reuse, efficiency, scalability and the characteristics of independence. We achieved good reasoning results by using this system.