Design And Construction Of DSB System For Bank Based On Data Warehouse

Nowadays the commercial Banks are facing the flood-like internet financial impact. As a result it is necessary to improve their IT capability. By a first-class processing of daily business operation information, banks can promote the management level and improve the risk control ability in order to adapt to the ever-changing market in this new era. Due to the complexity of correlation between banking systems, different types of data are distributed in a wide range of business systems without a union. In order to improve the efficiency of data analysis and raise the possibility of valuable information, banks have to learn more effective management skills and to uniform a huge amount of data between various business systems and operating procedures. It is significant to promote the scientific supervision and business process, so as to enhance the competence of the bank. At present, data warehouse platform can deal with these difficulties such as data islands between structures, the processing of the data from upstream system, and finally providing the data models to downstream systems. Therefore data warehouse is an incredibly useful tool which can offer an effective data basis to the bank management board making strategic decisions.In this paper, I will sketch the basic concept of data warehousing and its ETL (Extraction Transformation and Loading) process, and will also describe the research history by studying several domestic and overseas cases. At the same time, a comprehensive analysis of the procedure of the DSB (Data service bus from a real case) will be presented as well. Based on the existing technologies of data warehouse architecture, my study is further combined with the specific needs of upstream and downstream systems by offering data support and building certain data warehouse models. In addition, I will divide the types of data from the ETL schema into several levels such as SDM (Source Data Model), FDM (Foundational Data Model), MDM (Market Data Model) and RDM (Report Data Model). Furthermore, I will present a feasibility plan of ETL schema, which contains the entire ETL process including the shell scripts. Also this paper will introduce the ordinary data warehouse tools, especially the ETL automatic running tools of general banking industry. A number of important downstream business systems related to DSB and their running situations will then be briefly introduced as well. In the end, I will also introduce the functional testing and the performance testing which I designed after DSB system development.