Application Research On Information System Planning Based On CBR

The construction of information system needs scientific planning and right guidance, because it is a large investment, long period and technically complex system engineering. To improve the scientificity, information system planning (ISP) must be effectively controlled and evaluated. In this paper, Case-Based Reasoning (CBR) is applied in the process of ISP evaluation and a new ISP evaluation method based on case inference is proposed.First, this paper analyzes ISP and its evaluation research problems, and it reviews theoretical bases and key technologies of CBR. Second, the paper applies the case-based reasoning technology to the whole process of ISP evaluation. Then the case representation model and case retrieval method are explored. With respect to case representation, the XML-based representation method and relational database technology are employed to describe and organize cases. In the case retrieval, it is conducted by the hybrid retrieval method combined with knowledge-guided and nearest-neighbor approach. Due to the complexity of ISP evaluation, this thesis adopts a method of multi-layer grey relational analysis and weighted Euclidean distance in the similarity calculation to guarantee the quality of case retrieval. Finally, the paper conducts the ISP evaluation retrieval experiment by organizing the cases base.