Research Of The Application Of A Bank Embedded Data Mining

With the development of information technology and popularization of computer application, more and more electronic data has been produced. A "data rich and poor knowledge" problem has appeared when people in the face of huge amounts of data. In the1980s, the rise of data mining technology and database knowledge discovery technology opened the way to solve the problem. After30years of development, Data mining technology has been widely used in Data analysis and knowledge discovery from all walks of life. However, there are still many problems with currently data mining technology:(1) data mining systems provide only isolated knowledge discovery, it is difficult to embed large application systems;(2) the data mining engine are loosely coupled with database or data warehouse systems;(3) data mining industry is highly fragmented, companies and research institutions develop their own independent data mining systems and platforms, there is no open standards, can’t provide cross-platform support, can’t provide across the data warehouse system support. So it spends a lot of time to load and convert data in the process of data mining, and the algorithm running time is long, the efficiency is low. In order to facilitate the application of data mining, designing and developing data mining system has become a current research hot spot.From the theoretical perspective, introduced the history and current research data mining system, explored the entire process of data mining analysis and implementation steps, and then further studied the relationship between data mining system and a database or a data warehouse based on the existing data mining, proposed the view of embedded a database into data mining system, so that the data mining system tightly coupled with database system. Analyzed the cause of embedding Derby database into data mining systems, then embedded Derby database into data mining system in the form of Eclipse plugin, designed of embedded data mining system.From the application point of view, selected A bank credit card business as an analytical background, researched and analyzed A Bank Credit Card Center organizational condition and credit risk management methods to determine problems in A bank credit card risk management business, analyzed key factors affecting A bank’s credit risk, did association analysis about customer behavior patterns, then applied embedded data mining technology into credit card risk management business innovatively. Optimize the traditional Apriori algorithm, in view of the embedded data mining application in the management of credit risk, using two kinds of typical data mining algorithm in data mining association rules on the embedded data mining system, to provide decision-making reference for A Bank in credit card risk management business, reduce the credit fraud, and verify the Apriori algorithm of optimization on the embedded data mining system efficiency.In view of the above problem, this article from the related basic theory research of the embedded data mining technology, using A bank’s credit card risk management business data mining system for stimulating thinking. This paper further studied the embedded data mining system from business analytics, data mining process analysis, system requirements analysis, system design and other aspects.