Study And Application Of Data Mining In The Teaching Website

The purpose of building a teaching website is to provide a powerful platform for interaction between teaching and studying. As an important method of teaching assistance, teaching website has a vigorous development already. However, the existing teaching websites have lots of defects. In structure, they have much redundant building, bad code reusability, and adhesion promotion difficult and so on. In content, they cannot form a very good organization of knowledge and unify personal characteristic to form dynamic teaching. That has caused two problems. On one hand, the maintenance staff cannot display the site's function very well. On the other hand, the website content cannot arouse visitor's enthusiasm. Thus website will lose its construction significance. Along with the further construction of university's multimedia and network resource, how to develop an effective and flexible studying website according to the need of education becomes the main topic of this research.This paper is divided into six parts. The first part gives the concept of "teaching website system", makes an essential summary to the present situation of teaching website research, and proposes the main significance of this topic. The second part is about "Key Technology" and "Core Theory". It introduces the technology of struts and ibatis used in the construction of website and k-means arithmetic used in data mining. The third part, against the defects of k-means such as dependence of k value and initial center, proposes a new algorithm called KDM. Compare of purity and cohesion (tightness and separation) proved that the new algorithm achieves better result. The fourth part puts the improvement algorithm into a teaching website and provides a data processing technique. The fifth part introduces the detailed Struts framework, and proposes a synthetically framework based on struts and ibatis, which can strengthen the website's reusability and maintainability. At last, the paper gets the conclusion and the shortage of the algorithm to be solved in the future.