Research On Modeling Of Knowledge Based System In Application Area

Knowledge Engineering study the general expression and methods aiming at organizing knowledge in computer, setting up high quality knowledge bases , seeking useful knowledge and dealing problems using the knowledge. Nowadays, the technology of information processing which computer is used as tools went through numerical calculation stage, numerical processing stage, coming into the knowledge processing stage. Knowledge Engineering is now a key technology in the upcoming knowledge society and has become a major technique for information integration. This thesis systematically investigates the system composing element, expression, controlling, sharing and modeling of knowledge based system on the basis of analyzing theory developing status and the basic characteristics of knowledge engineering domain and the shortcoming that the knowledge based system modeling framework faces. The main findings of the dissertation are summarized as follows:(1) The V model design ideology of system engineering was implicated in the requirement decomposing of knowledge based system. The intelligent system is divided into system requirement(requirement ontology), domain conception knowledge(conception ontology), system knowledge(method ontology), control knowledge and explanatory knowledge corresponding to user's view, domain conception view, system resolving approach view, system structure view and system resolving track view. The data structure of these elements and the relationship among each other are definitude.(2) A hierarchical system planning scheme is firstly put out based on hierarchical case-based planning(HCBP) for solving the requirement decomposition of complicated knowledge based system. The high level of planning is for laying out a working outline, and the low level of planning is mapped to conception ontology and method ontology for reasoning and implementation. So, it is efficient and the basic knowledge chunks can be reused. A state and object based similarity metric method is proposed, and the process of decomposing, integrating and studying of HCBP is also given.(3) The realizing strategy of ontology maintaining and management are discussed based on knowledge consistency checking and ontology maintaining mechanism. The architecture of the computer system is given. One ontology conception classifies and study arithmetic is proposed based on decision tree principle.(4) In the domain of ontology sharing and alternation, the syntax heterogeneity and semantic heterogeneity of ontology were mainly studied, a hybrid system scheme which combine single global ontology and multiple local ontology's was proposed, and the semantic mapping between them was analyzed. RDFS are expanded and renewed aiming at the characteristic of thesis data element on the bases of the OIL's extending to RDFS. The syntax mapping between conception ontology and OIL's, APPL's extending to RDFS is established. The conversion method between conception ontology and XML is also given.(5) The knowledge based system control framework is proposed which the goal oriented blackboard control model provide the backbone and the ontology agents act as various domainrepresentative models. The structure and ameliorate arithmetic of the goal oriented blackboard control model were put out based on HCBP.(6) The methodology of knowledge based system developing process is also discussed. The system strategies of knowledge engineering are systematically presented by using system engineering methodology. The developing process are divided into knowledge identify stage, ontology model stage, knowledge extract stage, system implement stage, test, validate and modify stage etc. the content of every stage are definitude.(7) The agriculture crop schema decision support system of Qindu District of Xianyang city are built by applying the production of this thesis.