Research On CBR-Based Intelligent Designing System Of EC Website

Because of the development of network, E-commerce is now used widely, and it is playing a more and more important part in enterprises' development. However, it is still difficult for medium and small enterprises to develop themselves EC, e.g. the construction of EC Website, for it is too costly and time-consuming for them to build a complete and advanced commercial website system. Enterprises have to wait for several months only for a thousands-cost website and what is more headache is the cost of the sustain fee. All of that retrieves the development of EC. To solve these problems, this paper presents the idea of using the CBR theory into the designing of EC web. It designs an expert system based on CBR—Intelligent Designing System of EC Website (IDSEB). in order to realize the automation in the building process of EC websites. This system can produce optimized designing schemes based on enterprises' practical status and their requirements and it can improve the websites when their users change. In this system, the CBR methods of AI are used because there are some certain fixed schemas in EC websites. CBR is a newly developed methodology in the field of AI and its basic idea is to solve current problems using past practical cases. Based on this kind of direction, we build our case base using some successful EC designing schemes. When a new EC website is needed, the case reasoning procedure searches a proper designing scheme according to the requirement of users, and then the designers complete it to get the ultimate designing scheme. Based on the scheme, developers develop the website by the requirement of users and get the personalized EC website. The essential problems of IDSEB designing and development are knowledge representation of EC website cases and designing of the corresponding reasoning algorithms. To represent knowledge, the paper describes the structure and key content of EC websites using UML expendable mechanism, and then describes the UML structure using XML, and stores the ultimate XML documents as cases. The paper defines an algorithm of measuring similar degree between different case properties by their values. As for case searching algorithms, this paper presents an improved K-D tree algorithm based on the characteristics of XML documents and K-D tree algorithms and implements the algorithm. According to the above ideas, the paper presents a proper solution scheme of based-on IDSEB...