Clickstream Data Warehouse Based On Sql Server Analysis Service

Data warehouse technology originated from requirement of analysis numerous data .under sharply competitive enviroment for market.OLAP that is used as decision analysis based on business data is playing more and more important roles. Traditional OLTP emphase on updating database ,which add information into database,while OLAP is acquiring and using information in database. Hence, Raphl Kimball ,famouse data warehouse exerts, said that we had spend twenty years to put data into database, it is time that we fetch them from database. Click stream data warehouse is one of devoloping direction of Data warehouse technology. With quick development of E-commerce,E-commerce has exerted huge impact on enterprice's running. E-commerce increasingly become a improtant channel of marketing and customer serving. Fully Using E-commerce will bring greate economic profit and social profits to enterprise. E-commerce site make a great deal click stream data ever day ,which include many useful information. For example, where are customer from, what action customer took, what did customer interest in. analysising these data can not only direct building of site, strengthen adhesion, but reflect enterprise'status on marketing and finace. In short,deeply analysising these data may help owner of site to improve customer relationship and qualities of service. In this thesis ,author research dimentional modling method of date warehouse, ETL design of click stream data warehouse,implemention of Data warehouse and front-end displaying technology. In dimentional modling, author design data warehouse by method that suggested by Raphl Kimball. In ETL design of data warehouse,author firstly summarize ETL methods . because the click stream Data warehouse is based on SQL Server 2000, DTS is choiced as tool to design the Data warehouse's ETL . DTS has strong functions and may be used easily. Especially, for the date warehouse built by using SQL Server 2000, DTL is ideal tools for ETL design. After built the data warehouse,mutidimetional dataset are needed in order to take OLAP analysis. Dataset is data set that is build from Data warehouse and is orgnised and summarised into a CUBE. In the project, author choice Sql Server Analysis Service to design multidimetional dataset. Its Analysis Service is a middle server used for OLAP and data mining. Analysis Server orgnise data that are summurised-computed at advanced into multidimentional dataset. In design of front-end displaying, author choiced Pivot-Table to display mutidimentional dataset. Pivot-table is client end components of Analisys Service. User may access data in multidimentional dataset by it. In design of front-end displaying,MDX is used to take theme-orinted analysis to multidimentional dataset.then result are displayed in web pages.