Abstract: With the rapid growth of the Web data, for the phenomenon of data explosion but poor knowledge, as the data mining based on database table has been unable to meet the needs of Web application, a distributed Web mining is becoming a research hot spot presently. Firstly, based on the research of web mining, XML is used to transform semi-structured data to well-structured data, and a distributed web mining model based on XML is deeply discussed. Then, a mobile agent technology is introduced into the design of Web mining system, and a distributed Web mining structure based on data mining agent is built. Finally, an architecture design of network teaching platform based on XML and agents is realized, which can provide intelligent aids for personalized learning, thus greatly improve the teaching efficiency of network teaching platform. Web data mining is a new important research field in data mining. In this paper, the conception and characteristic of data mining based on Web are introduced the process and the general methods of data mining based on Web are expatiated. At present many websites are built with HTML, which is difficult to achieve real effective and accurate web mining. The appearance of XML has brought convenience for it. Based on the research of web mining and a model of web mining system which has basic data mining function and faces multi-data on the Web is built. With the development of information technologies, web data mining has been put forward and in wide research. It is defined as the discovery, extraction and analysis of useful and potential information from the World Wide Web. But much of inhomogeneous and anomalistic and dynamic updated semi-structured data in web pages makes web data mining difficult. To solve this problem, on the basis of analyzing the characteristics of XML, the paper presents a web data mining model on XML, introduces the method to implement the model with XML.

Keywords: XML Database, ASP.net, SQL server.