A vehicle insurance risk assessment method, device, equipment and storage medium

By using federated learning to extract and fuse encrypted features from vehicle and insurance data nodes, the problem of cross-source information utilization and privacy security of vehicle network and insurance data is solved, and efficient and secure vehicle insurance risk assessment is achieved.

CN122243656APending Publication Date: 2026-06-19TIANJIN FAW TOYOTA MOTOR CO LTD
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
TIANJIN FAW TOYOTA MOTOR CO LTD
Filing Date
2026-03-23
Publication Date
2026-06-19

AI Technical Summary

⚠Technical Problem

In the integration of vehicle-to-everything (V2X) technology with insurance business, raw data faces the risk of leakage and is difficult to effectively aggregate within the same computing node, resulting in insufficient scoring accuracy and timeliness.

⚗Method used

The federated learning approach is adopted, which uses a feature extraction model with the same initial structure and co-trained to extract encrypted features from the data at the vehicle data node and the insurance data node, respectively. The data is then fused and periodically updated at the evaluation node to generate a fused feature vector for risk assessment.

🎯Benefits of technology

It enables the utilization of information across data sources, ensuring data privacy and security while maintaining the timeliness and accuracy of scoring results, thus solving the problems of data leakage risk and insufficient scoring accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

This application provides a method, apparatus, device, and storage medium for vehicle insurance risk assessment, relating to the field of information technology. The method includes: receiving a first encrypted feature vector from a vehicle data node and a second encrypted feature vector from an insurance data node; fusing the first and second encrypted feature vectors to generate a fused feature vector; and inputting the fused feature vector into a risk assessment model to obtain a corresponding vehicle insurance risk score. The risk assessment model is a model trained based on sample fused feature vectors and corresponding risk labels, used to map the fused feature vectors to risk scores. The first feature extraction model, the second feature extraction model, and the risk assessment model are configured to periodically and collaboratively update based on newly added original vehicle data, original insurance data, and corresponding risk labels.
Need to check novelty before this filing date? Find Prior Art