Research And Application Of Web Usage Mining In Website Optimization

Web Usage mining is one of the most important research directions in Web mining. It mainly focuses on Web usage data. From the point of view of server, the object of the Web usage mining is the valuable information provided by the Web server. The conclusion of Web usage mining could be used for Web designing. From the point of view of user click sequence, user's information is monitored and could be used for previously storing of Web pages.The thesis is mainly to transform the Web usage data from the text status to format status at the first place then a model of click stream warehouse is rebuild based on the model given by Mark Sweiger and Mark R.Madsen and applied to Website after the data of a Website is imported into the database. At last we analyze the data by the tool provided by the SQL 2000 server analysis and based on the conclusions of the analysis we provide suggestions to the administrator regarding the structure of the Website.The thesis mainly focuses on three aspects as following:Firstly, the thesis offers effective algorithms for data import, user recognization and session recognization. The Web usage data refined through these algorithms would guarantee the data source for the data warehouse and data mining.Secondly, the standard click stream warehouse model is customized and adapted to the Website instance. Then the logical model is applied in the database, which contribute to the setup of the physical model. At last the data of the Website is transported into respective data tables that provide the data source for data mining.Lastly the algorithms and the model mentioned above are applied into the data collected from "myjob" website over two months. Then the formatted data will be analyzed through the decision tree model that provides different views of data for analyzer. The conclusions is extracted from the data view by analyzer and applied into the reconstruction of the website.