The Design And Application Of ETL In Auditing Systems

In financial field, especially commercial bank, users put forward high request to the data real-time property. Among some systems that bank use, such as auditing systems, users need to make statistics and analyses on transaction data in a very short time. Data used by systems need to arrive in a short time, but data sources are various. Auditors require all-round information, which contain not only personal loan service and enterprise credit business, but also international business, capital service and so on. These services have their own systems, and some data are from core system. Moreover, data are stored in heterogeneous environment. For instance, different databases, different operating systems are used, etc. It is a great challenge for us to extract data from numerous systems and form a data market needed by auditing systems rapidly.According to the statement above, this paper will focus on the study of ETL(Data Extract, Data Transform, Data Load) in auditing systems. The meaning of ETL is data extraction, transformation and loading. The importance and necessity of ETL were analyzed in the audit system of commercial bank, and the ETL's design and application were completed by the system's demand and business environment. An ETL model based on business is designed. In comparison with traditional, mature, commercial ETL tools, partial implementation of this model are completed by means of codes directly, which makes more pertinent and flexible. It can process data transformation and loading tasks in complex systems of bank. Most importantly, it also could cut some unnecessary execution steps in commercial tools to reduce the time. To compete with other financial system's models, this model is designed and realized based on business, and possesses better error recovery mechanism. Through the system's ETL task attemper, we achieved the data extract,data transform and data load form business data to the target data market. Practical application shows that this ETL model meet demands well. The data is integration an unified. It helps the staff get high-quality data quickly, and it also reduced the workload of staff. Therefore the efficiency of audit work could be improved.