The Study Of The Technology Gathering Information From The Internet

Mining large-scale data on the Internet and recommending interesting information for users intelligently are the research emphases home and abroad in last years. Web Usage Mining is the method of having better application value.This dissertation includes mainly four parts: The first one introduces the definition, basic process, classification, main methods and general technologies of the Data Mining. The second one summarizes the definition, process, classification and basic thoughts and key technologies of three sorts of Web Mining, and compares it with Web Information Index technology. The third one studies the method of the data preparation and the algorithm of mining frequent browsing path based on maximal forward reference in the course of the Web Usage Mining. The last one discusses simply the method of expressing the different and half-structural Web data into structural data with XML.The innovation of this dissertation is: It attempts to design a intelligent Web site system model based on Web Usage Mining and the algorithm of pre-judging and recommending. The system consists of four parts: management interface, preparation module, mining browsing patterns module and intelligent recommendation module. The main function is: it inspects the browsing behavior of user on-line, recommends variably the most relevant interesting pages according to the frequent traversal paths from Web Usage Mining and last sends them to the cache of user in advance. It saves the browsing time and expenditure of users. The scheme completes not only to gather information intelligently but also to enhance system performance. Users do not know the process of mining, recommending and pre-sending at all. The system has better adaptability by means of flexible parameter setting and control using management interface.