| After years' construction of information technology, the Business systems and departmental management information systems have accumulated a lot of data of business and customers. However, these data are distributed in independent systems and servers of the bank distributed, and mostly based on confusion, resulting in too many redundant data, the inconsistencies between data and data, and the low data quality. If these data can be used for bank's management and decision support, it will produce more significant benefits. This thesis using commercial bank product management system as an example, focuses on the ODS data warehouse research in the bank application.This thesis, through comparing data warehouse and ODS on the basis of theoretical knowledge of them, explains the role and advantages of the ODS. For the situation of the storage in bank and the method of DW construction, this paper gives a three-tier ODS data warehouse architecture, those are data storage layer, data acquisition layer and data display layer. According to the application samples of the product management system, this paper through the system requirements analysis gives the overall design, and make the detailed design of data storage layer, data acquisition layer and data display layer one by one, that are the three-tier architecture of ODS data warehouse. model designing for the standard data layer, basic/summary data layer and data mart layer is made to complete the design of the data storage layer. The ETL process based on storage procedure is made to complete the implement of the data acquisition layer. By SUP and Cognos I complete the implement of the data display layer.At present the system has been officially running, it resolves the problem of the dispersion of products and business data, provides staffs with a product management and analysis platform, and has played a positive role for the planning of bank's internal data and decision support. |
PDF Full Text Request