Large Commercial Banks In Enterprise-class Data Warehouse System Build

Currently in China, the commercial banking sector is in the process of fast evolution. For its huge data size, complex banking techniques, the high quality requirements on risk control management, and pressing need of data consolidation, the technique of data warehouse is urgently required. The applications of data warehouse are playing more and more big and essential role in the process of analysis, prediction, deducing and decision-making, and are also bringing considerable profit for enterprises.So far there are various overseas banks which had constructed enterprise data warehouse for years. They are strongly experienced and get a lot of attachments on it. Through enterprise data warehouse, the banks are able to consolidate all of the distribute data in OLTP and data marts to build a complete data view. It will be great helpful to provide complete, efficient and accurate information for downstream applications.The thesis analyzes the features of banking operations and data of large commercial bank, combining with the implementation practice in big commercial banks, to comprehensively study and describe the enterprise data warehouse applications in commercial banks, and the specific design and implementation of each module. The whole enterprise data warehouse system is mainly composed of the modules including data model, ETL data loading, metadata management, data quality management, data service, data backup and restore. Each module had been optimized in designing and developing based on the data warehouse designing concepts and the characteristics of commercial banks to ensure the ease of use, efficiency, accuracy and security.The characteristics of this system mainly lie in:integration and optimization for the large amount of data from source systems; designed an integrated, subject oriented data model for complex banking operations; designed parallel loading programs to improve the loading efficiency significantly; designed various loading algorithms to meet different data needs; enhanced metadata and data quality management to maintain the sustained and efficient operation of system; designed specific program for putting into production to minimize the impact to current systems. Currently the enterprise data warehouse system had been successfully put into production and is operating stably. The follow-up developing is also in steady progress. Enterprise data warehouse represents the requirement and the evolution direction of commercial baking data management to resolve the issue of integrity for data, meanwhile it provides good data support for a variety of business units to make customer service and decision making better and accurate. The effect of providing impetus for developing banking sector will become more and more apparent.