| Currently building centralized data warehouse for enterprise scale systems is one of the major directions for every large commercial bank. Through data warehouse we can effectively improve financial enterprise management over risk, customer relationship, sales and marketing. The overall quality of data warehouse projects depends nearly on the quality of data from different data sources and the processing toward the data. That is to say, data quality problem is closely related to the customer's trust and confidence in data warehouse.Focusing on the comprehensive utilization of financial data in certain commercial bank, this dissertation did some correlative studies of data quality management based on the database of Teradata, Control-M workflow platform and applied the method of AOP and data mining. The main content is as follow:1) This dissertation reviewed the recent developments of data quality at home and abroad. Then, it described the concept of data warehouse, data quality and the application of ETL method in data quality assurance. Finally, it elucidated the relationship between software architecture, metadata and data quality.2) This dissertation added the domain knowledge and quality factors to realize the quality-driven data warehouse.3) This dissertation proposed a method that combining the idea of RBAC and AOP to control the privilege of data quality manager, solved the complex role distribution in the data warehouse management, applied workflow schedule to ETL in data warehouse.4) This dissertation combined with concrete situation in a certain commercial bank, analyzed the reasons resulting in data quality problem and designed a metadata management system for the data warehouse, furthermore proposed the management process of data quality and the process of quality inspection which is practiced and verified. The result illustrated that the implement of this method improved the data quality of data warehouse and the customer satisfaction. |
PDF Full Text Request