Credit Card Risk Assessment Application of the Ministry of Human Resources and Social Security — First External Query Service for Social Security Big Data

Since 2016, the Information Center of the Ministry of Human Resources and Social Security has leveraged networked monitoring data from cardholder databases, employment, and social insurance to study the social security credit evaluation model, forming a c

Value Delivered

l  Innovative Application of Big Data in Human Resources and Social Security:Provided personal credit information to the banking system for the first time through a credit scoring query system, truly realizing data monetization.


l  Multi-Cluster Application:Actively explored using GBase 8a, moving from a single cluster to multi-cluster deployment. Customers mastered data interaction methods between clusters and gained deeper product insights, laying the foundation for the continued deepening of GBase 8a applications at the Ministry of Human Resources and Social Security.


l  High-Performance Querying:Met the requirement of precise queries on over one billion data records returning results within one second.

Solutions

The Ministry of Human Resources and Social Security (MOHRSS) originally had a 10-node GBase 8a MPP Cluster production cluster, primarily storing social insurance, employment, and other information. This new credit scoring service leverages existing data in the production cluster to conduct credit evaluation model research and establish a scorecard model. Individual credit scores are calculated using factors such as age and social insurance status. The scoring results are then transmitted to a credit scoring query cluster (a newly built 3-node cluster). This query cluster provides external query and display services and is offered to a bank via a web service interface.



The credit scoring query application and cluster must provide data query services over the external network, so the credit scoring query cluster is deployed in the external network, while the main production cluster is deployed in the internal network. Network policies are configured between the two clusters to ensure data security between the internal and external networks. Data is transferred between the main cluster and the credit scoring query cluster via DBLink. Bank query clients are restricted by IP and connect to the MOHRSS credit query server through a dedicated line.

Requirements Analysis

The credit card risk assessment scoring query application system, based on MOHRSS (Ministry of Human Resources and Social Security) data, is designed to handle both internal and external queries of MOHRSS credit scores. The overall objective is to enable MOHRSS credit scoring queries and provide query services for MOHRSS credit scoring information to both internal and external users. Core functions include scoring data import, query permission and authorization management, and query interface management.


Supports up to 70 concurrent clients, with query results returned within 1 second. The source dataset contains hundreds of millions of records.

Project Background

Since 2016, the Information Center of the Ministry of Human Resources and Social Security (MOHRSS) has been researching social credit evaluation models using data from the national cardholder database, employment, and social insurance monitoring systems. This led to the development of a social credit evaluation index system and model. To better support innovative applications of this social credit framework within the broader societal credit system, a pilot project was launched to explore credit card risk assessment based on human resources and social security data. The immediate task is to build a credit card risk assessment scoring query application system that leverages this data to generate and retrieve credit scores.