Research On Design And Implementation Of Credit Risk Assessment System In Rural Credit Cooperatives

With China’s accession to the WTO, the whole world continues to accelerate theprocess of economic integration. The competition among different domestic majorbanks has been intensifying. Besides, the competition between domestic banks andmajor international banks has reached a boiling point. Strengthening modernmanagement of banks and making scientific decisions are the core weapons for banksparticipating in the competitions. Currently, IT technology is advancing at a rapidspeed, so that a lot of banks begin to look into computer technology and internettechnology for help. It is the key points for the banks to improve their corecompetitiveness that building an intelligent decision-making through computertechnology. For the reason that banks are able to get potential information effectivelyand efficiently, make an in-depth analysis of the banks’ credit risk in order tominimize the credit risk of bank loans.This study of loan risk evaluation system of rural credit cooperatives does meetthe "New Basel Capital Accord" and" credit risk five Classification" requirements,using artificial intelligence expert system and inference engine technology. Accordingto the credit risk assessment standards of rural credit cooperatives, this essay is basedon java platform to develop the customer loans risk assessment system, which suitsthe characteristics of China’s rural credit cooperatives. The system is powerful andcan automatically correct and update the database, knowledge base and credit modellibrary. In the bottom of the system, with the inference engine and two BP neuralnetwork inference (rule-based reasoning and case-based reasoning), we firstly studythe loan risk evaluation model of both firms and individuals concerning qualitativeand quantitative indicators. Then, the system automatically generates the corporatecredit rating of loan customers. In this way, banks can get more reference about riskcontrol to reduce the risk of loans.This paper focuses on expert system’s weaknesses, delving into the credit model,knowledge base, inference mechanisms, and building farmers risk assessment modelwhich is based on BP neural networks and the expert system. Also, it has yielded goodresults in the part of practical application. The introduction of BP neural network canweaken the human factors of the weight determination and improve the accuracy andauthority of assessment results. To some extent, the model successfully enhance the system reasoning efficiency and ensure the stability of the system.In the end of the study, this paper rethinks and summarizes the entire process ofthe study. We conclude there are problems in this system, but it has laid thefoundation for the further refinement in the future.