Research And Application Of Data Warehouse,Data Mining In Risk-Control Of Credit Cooperative

Since China joined the WTO,with the increasing opening of finance service market,banks in our country are involved in gradually changing challenge,and the industrial competition is becoming more and more fierce.The competition form of finance service market has been evoluted gradually from the enterprise's competition to the competition of Business Intelligence (BI).Now, the rural credit cooperative is finishing the collection of basic data in each province,so it is very important and urgent to build an advanced analysis system based data warehouse(DW),olap and data mining(DM) in this new situation.The new build system can help bank manager make right decision which will enhance the credit cooperative's core competition ability.The rural credit cooperative is special in finance industry ,so there are many strong generalization and adjustability in every location application.Jiangsu Haian credit cooperatives have already completed the data collection, but they haven't have a intelligence analysis system based DW.As credit cooperative's financial products are relatively single,main business of current is absorbing deposit and providing loan. So risk control about loan is very important.Associated with credit cooperative of Jiangsu Haian's business, the paper mainly study the risk control analysis system(XYSDW) adapted in credit cooperative,which focuses on credit loan operation and digs rules related with credit loan risk. the main researches are as follows:(1) Provided a DW structure adapted to credit cooperative—XYSDW,which using Microsoft's SSIS as the ETL tools.Learn the relevant knowledge and make a research on credit loan risk to meet the analysis demand.Build the datawarehouse flatform and multidimensional data model by using Microsoft's SSAS.Using Microsoft's SSRS present the analysis application of client.(2) Choose ID3 decision tree algorithm to dig credit loan operational data,and present the double support concept.The ID3 algorithm was improved by using double support concept,which can make a decision to a new loan throught mining the old credit loan operational data.(3) Study the sequence model data mining,present a improved GSP algorithm in SQL.Use it dig risk evolvement rules in credit loan risk sequence database.And predict the customer's next possible situation grade,which can help creditor and risk control manager to make some relevant strategy at first time. Lastly, the problems existed presently are summarized,and the expectation towards future research is given at the end.