基于机器学习的保险数据处理方法、装置及存储介质
By using machine learning methods to mine insurance feature behavior representation vectors from insurance policy claims datasets, the data application barriers in credit risk assessment have been overcome. This has enabled the efficient extraction of insurance features that are highly coupled with credit risk, thereby improving the accuracy and coverage of credit risk assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BANK OF CHINA INSURANCE INFORMATION TECH MANAGEMENT
- Filing Date
- 2025-10-15
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional credit risk assessment faces problems such as lack of transparency, outdated information, and high verification costs when serving small and micro enterprises. It is difficult to dynamically reflect the true operating conditions and potential risks of enterprises. Furthermore, there are barriers to the application of insurance data in the credit field, such as the complexity of building an accurate correlation system, significant data noise, and differences in cross-industry standards.
By using machine learning methods, we can mine insurance characteristic behavior representation vectors from insurance policy claims data. We can then use the commonality coefficient and co-occurrence coefficient of the representation vectors to adjust and extract highly correlated enterprise feature labels, forming a reliable risk profile and improving the accuracy and coverage of credit risk assessment.
It enables in-depth mining of inherent correlations from massive insurance data, improves the accuracy of feature extraction, ensures consistency between insurance and credit businesses in financial enterprises, and enhances the adaptability of risk control efficiency and financial service coverage.
Smart Images
Figure CN121304353B_ABST