Anti-fraud and credit risk prediction method and system based on association network

Abstract

The invention relates to an anti-fraud and credit risk prediction method and system based on an association network. The method comprises: determining whether a tested user is a fraud user listed in ablacklist or not; if the tested user is not the fraudulent user, obtaining a relationship group having data interaction with the tested user, and constructing an association network through the relationship group; and analyzing node objects in the association network, and determining a credit score of the tested user so as to carry out risk early warning. According to the method, the credit scoreof the tested user can be determined more accurately by constructing the association network of the tested user and predicting the credit risk of the tested user through the association network, theprediction accuracy is improved, risk early warning is carried out when the credit score does not reach the standard, risk avoidance is facilitated, and the economic benefits of banks and other financial institutions are ensured.

Description

technical field [0001] The invention relates to the technical field of financial risk assessment, in particular to an anti-fraud and credit risk prediction method and system based on an associated network. Background technique [0002] With the development of the economy, the financial industry is becoming more and more developed. Among them, loans are an important part of the financial business. The types of loan business are becoming more and more prosperous, and the loan channels are becoming more and more abundant. In people's life and production activities, ranging from personal daily consumption to large-scale production and operation of enterprises, loan business is inseparable. Financial institutions that provide loan services generally conduct a credit risk assessment of the loan business before granting loans, and only when the risk meets the requirements will the loan application be approved. Most of the existing credit risk prediction methods first confirm the i...

AI Technical Summary

Problems solved by technology

[0003] Although the existing credit risk prediction methods are various, there are still many incidents of fraudulent bank loans in real life, which bring economic losses to banks.

Existing credit risk prediction methods only examine the personal status of the loan applicant, and do not fully consider other people who have mutual interests with the applicant. For example, the loan applicant has a very frequent contact, and the contact has is a fraudster, in which case the applicant is likely to be influenced by the contact to commit fraud

In addition, the existing credit risk prediction method only conducts credit risk assessment before the loan application is approved, but after the application is approved, the credit risk of the lender will not be tracked and predicted

Especially for businesses that require multiple loans, such as credit cards, the bank only predicts the credit risk of the applicant at the credit card application stage. Once the credit card is issued, the user can use it arbitrarily within the limit, which has hidden dangers, such as If it is misused by others, or within the validity period of the credit card, the cardholder has a major problem that will affect his credit, such as fraud. The issuing bank of the credit card will not conduct tracking credit evaluation for these behaviors. This defect may cause Cause economic loss to the cardholder or the bank

Images