Design And Implementation Of University Credit Test System Based On Knowledge Analysis

In order to make loan students who from Guizhou Province to learn the honesty repayment and reduce loans default,we developed "online answer system for student loans" according to the demand of Guizhou Province in 2012.The system was used to test students' loan knowledge.Relevant departments gave a simple assessment about their credit according to the students' examination results.But the system had problems such as poor scalability and code reusability,which can't meet the new needs of users.Therefore,this paper aims to redesign and realize the university honesty education examination system based on knowledge analysis.The system uses the decision tree classification algorithm to construct knowledge analysis model,which can find the potential correlation between examination scores and knowledge points.Combined with the loan default factor,this paper uses the BP neural network to establish a credit evaluation model to achieve the actual credit assessment of the loan students.The main results are as follows:(1)System design and implementation.Based on the limitation of the original system architecture,the key technologies such as WCF and NHibernate are used into the data service layer and data persistence layer to design a new system architecture,then designs the University Credit Education Examination System based on knowledge analysis.In the new system architecture,we can develop system for different terminals,and don't need to make any changes in database.(2)Knowledge analysis model's design and implementation.The examination can effectively detect the fraction of students to learn the knowledge acquisition.The test papers as the carrier of examination,which content is the examination of the knowledge points.So the analysis of examination paper should be based on knowledge analysis as the foothold.Therefore,this paper takes the examination paper examination knowledge as the foothold,using the decision tree classification technology to analysis the test paper,and constructing the knowledge analysis model.The paper takes the system paper as an example.We can obtain the gain rate of each knowledge point,which can know the students' situation based on decision tree classification between the goals and the knowledge points.The model can help developments to learn about the situation of students' master.(3)Credit evaluation model design and implementation.In order to support the relevant departments to understand the real credit students' loans,the paper applies BP neural network to design credit evaluation model according to scores,the annual income of the family,the school of historical default rates etc.We also use the loan data to take experiments.In addition,the simulation program was born as a DLL file,and the C# program in the reference DLL file to achieve this function to complete the loan credit assessment.Based on the ASP.NET,the system applies B/S and C# programming language,and combines with the above key technologies and theories to develop and implement.The system not only meets the needs of users,but also achieves the analysis of the test data.It also can effectively prevent the occurrence of default loans before the student lends.