The Research And Application Of The Data Mart Building Based On The Data Warehouse Technology

The technology of data warehouse is a combination of two computer technology,that is database and artificial intelligence. It is the mainstream of today's information management technology.Using the methods of artificial machine learning , knowledge processing and neural network method, it can mine useful information from a database, finding knowledge, and making clear the relationship among the numerous data in which it can find some regular patterns. It can facilitate the analysis for users, and also help users to store various information in database, thus decision-makers can make quick decisions, and produce great economic benefits for enterprises.Data mart is defined and designed to store messages in order to satisfy the demand of a department or a group of users. It is a batch data and business rules organized for a specific department of decision support. Compared with the data warehouse, the granularity of data mart is thick. So it is facilitating for the access analysis and quick searching and now it has bean used widely.This paper is based on the technology of data warehouse and data mart. With the audit business needs of PuFa bank, we built and designed the model of audit work platform data mart using Power Designer, and completed the extracting work from the data warehouse to the mid-database of ODS using IBM Date Stage ETL tools .By composing the stored procedure of Sybase, we transformed and loaded the data from ODS to the data mart. Furthermore, on the base of the well designed data mart, we have designed and implemented the individual bank business subsystem of the audit work platform with relational J2EE technology. Now the system has been online, and runs well.