The study of click-stream data warehouse

Click-stream refers to the information left by users who have browsed the website. Generally, the information is staying on the website server as web logs. By collecting the information of users transferring from page to page on the site we can keep track of visited linked nodes, including the source of the user location, routes of users browsing , which is of great significance to the analysis of users´ visiting habits. Therefore it is necessary to exact the knowledge from large scale data. This paper uses SQLServer2005 to finish the design and implement of a data warehouse, which depends on XX-international trade site web logs, and finally we get the law of users´ visiting. Data preprocessing, dimension settings, multidimensional data sets analysis and display of OLAP are in the final results of the paper.