Web Mining Technology And Its Applications

Data mining means extracting useful knowledge and information automatically from the mass of data. Data mining has become an important field in the study of database and machine learning. Currently World Wide Web is developing rapidly in broadness and depth. By applying the approaches of Data mining into web to solve some problems, a new field "Web mining" is presented.The objects of Web Mining include all kinds of Web data: content of web pages, structure between pages, usage information of users. With Data mining we can find useful knowledge, extract knowledge from WWW, improve web site designing, and develop e-commerce more effectively. In the process of building "Web usage information mining system", we have made a thorough study on the approaches of web mining, including: data cleaning, transaction recognization, clustering algorithm in web broadcasting, association rules discovery, and etc.The following is my main work:1. By data preprocessing, Non-structured information is organized into some transactions or sessions in a database. Then the web data can be processed by the classical methods of data mining. In addition, we can use data cleaning to exclude a lot of useless data, and improve the efficiency of mining activity2. We bring association rules discovery into web mining. Finding frequent itemset is the fundamental part of association rules discovery. We employ a rapid algorithm (Apriori algorithm), to produce the frequent itemset. By analyzing web usage information, we can find some rules of users' usage. Association rules discovery can be used in organizing web site, web broadcasting, and etc.3. When we broadcast Web pages through broad band broadcasting network, what to broadcast and how to broadcast are two problems. In this paper we provide anew web mining method (WebClustering) to solve these problems. This method combines the idea of cluster and association rules. The mining object is Web pages of Cache and Log in WWW Proxy server. By this method we can find a valuable Web broadcast set and create some index HTML pages to indicate the users to navigate.4. We introduce the general steps for Web usage mining, bring some key technique of data mining, such as data preprocessing, clustering algorithm, association rules, etc, into web usage mining. We give evaluation and verification to above technique by a series of experiments. Finally we built the prototype of a web usage mining system...