There is a large amount of customer registration data and related data, and the current problems of insufficient depth of data analysis and low coverage of rules have resulted in the failure to intercept some of the overdue people in the risk control business. It is necessary to find out the part of the population that cannot be intercepted by the current rules, so as to minimize the loss caused by overdue.
The consumer finance company, based on big data analysis of customer relationships and customer characteristic attribute data, mines rules for identifying customers at risk of default during the performance period (0-6 months) and outputs the group of individuals who may experience overdue payments in the future. The Wuji Intelligent Risk Identification Solution outputs rules and crowd packs to mine risky customer segments not recognized by the existing rule pool on the basis of existing mapping rules.
Inability to detect hidden rules manually, low rule coverage, successful granting of credit by gang fraud in credit operations, resulting in financial losses.
In the context of intensification, there are cross-scenario and cross-sector characteristics, and it is difficult to manually refine the rules of risk control.
Through the automatic discovery of wind control rules and deep data mining, the coverage of rules is effectively improved, and effective rules are constantly accumulated. Discover the core personnel of gang fraud, involving 32 customers, of which 14 were successfully granted credit.
The Wuji Intelligent Risk Identification solution supports multi-source data access and builds a multi-source data topology network, which provides a more comprehensive understanding of data relationships and improves rule accuracy.
This mining found 98 high risk correct rules and the rules identified 10,384 high risk lists.
High-risk customer hit rate increased from 40% to 70%.
The Affiliate Network intercepted 85.43% of affiliated gang fraud.
Grandhoo