Font Size: a A A

The Design And Implementation Of Personal Credit Evaluation System Based On Neo4j Database

Posted on:2020-07-19Degree:MasterType:Thesis
Country:ChinaCandidate:Q X DingFull Text:PDF
GTID:2518306518467744Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the development of credit business,the evaluation of users' personal credit has become more and more important.The personal credit evaluation system designed in this paper aims to provide reference index for enterprises to deal with credit business and reduce enterprise risk.This paper studies the development status of personal credit assessment technology at home and abroad,investigates the needs of enterprises for personal credit,and analyzes the functional and non-functional requirements of the system according to the business process and use case diagram.The architecture design of the system is completed,and the system is designed in outline and database.Established the use of the Spring framework,Neo4 j graphics database storage data,Java and Python language development background,Bootstrap and Ajax foreground technology.The system is mainly divided into user management module,data management module,personal credit model module,personal credit assessment module,application module of personal credit value and assessment scheme management module.For calculating initial credit personal values,in this paper,the BP neural network and logistic regression model,and serial combined the output of the BP neural network as a logistic regression model training with the original data field variables and independent variables and logistic regression model,according to the model to calculate the corresponding score card,and then calculate individual initial credit value.The calculation of individual weighted credit value is divided into two steps.Louvain algorithm based on parallel computing is designed to quickly divide the system users into communities.A linear threshold model based on Page Rank algorithm is proposed to quantify the influence of credit between different users.In actual business,business personnel can not only calculate the minimum personal credit value by using break-even method to quickly screen business applications,but also calculate the credit limit by using expert static method and linear interpolation method.The main functions of this system have been achieved and achieved the expected results,which can better assist the credit business staff in handling credit business.
Keywords/Search Tags:Neo4j, Personal Credit Evaluation, BP Neural Network, Logistic Regression, Louvain Algorithm, Linear Threshold Model
PDF Full Text Request
Related items