Click-stream Data Warehouse In Applied Research, E-commerce

E-commerce website makes a great deal of click stream data ever day. And with the increase of the network information in many fields, traditional database store can not satisfy clients'demand. In this condition, data warehouse has been springing up. The data warehouse is a kind of subject oriented, integrated, stable and time-changing dataset used to support the management in decision-making process. The difference between Data Warehouse and Data Base is that Data Base system focuses on transaction processing, but Data Warehouse system focuses on analysis. The data warehouse built for the click-stream analysis is known as click stream data warehouse. Click Stream Data Warehouse is a new application after the increasingly widespread of Internet, especially after the rapid development of e-commerce. With quick development of E-commerce,E-commerce has exerted huge impact on running enterprise. The application of Click Stream Data Warehouse can not only direct the building of E-commerce website, strengthen adhesion, but reflect enterprise's status on marketing and finance, and then may help owners of E-commerce website to improve the relationship with clients and qualities of service.This thesis will research dimensionality modeling method of Data Warehouse, ETL design of click stream data warehouse,implementation of Data warehouse and front-end displaying technology, and taking china-pub as the research object, this article adopts the unified approach of Dimensional modeling to build a prototype of Click Stream Data Warehouse based on Microsoft SQL Server 2005.We use IS (Integration Service) in ETL. After the building of the data warehouse database, the design of dimensional dataset is necessary in order to carry out the OLAP analysis, and this paper chooses Microsoft's SQL Server Analysis Service to design a multi-dimensional dataset. Its Analysis Service is a middle server used for OLAP and data mining. Analysis Service organizes data converged-computed in advance from the data warehouse to multi-dimensional dataset. At last Reporting forms are used for the website business analysis and user behavior analysis.