A data desensitization and feature extraction method based on user privacy protection

By using dynamic hierarchical desensitization and differentiated processing, combined with hardware security modules and deep learning models, the contradiction between user privacy protection and data utilization value in existing technologies is resolved. This achieves efficient data desensitization and feature extraction, and constructs a multi-level privacy defense system suitable for big data platforms and artificial intelligence model training scenarios.

CN121935969BActive Publication Date: 2026-07-21GUANGDONG GUOLI EDUCATION TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGDONG GUOLI EDUCATION TECH CO LTD
Filing Date
2026-03-24
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing data anonymization technologies suffer from issues such as feature loss, a conflict between security and performance, fragile key management, and limited privacy protection, failing to effectively protect user privacy and impacting the value of data utilization.

Method used

It employs dynamic hierarchical desensitization, differentiated desensitization algorithms, hardware security module key management, multi-level feature extraction framework, and differential privacy secondary protection, combined with LSTM network, GeoHash encoding, and BERT model, to achieve differentiated processing and feature preservation of sensitive data.

Benefits of technology

It achieves the protection of user privacy while preserving the statistical characteristics of data, enhancing the value of data utilization, meeting the real-time processing needs in high-concurrency scenarios, and building a multi-level privacy defense system to resist link attacks and inference attacks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121935969B_ABST
    Figure CN121935969B_ABST
Patent Text Reader

Abstract

The application provides a data desensitization and feature extraction method based on user privacy protection, comprising a sensitive data automatic grading step, a desensitization processing step, a feature extraction step and a privacy risk verification step. Through mechanisms such as dynamic grading desensitization, differential desensitization algorithm, hardware security module key management, multi-level feature extraction framework, differential privacy secondary protection and privacy-utility evaluation model, the application can solve the problems of feature loss, security and performance contradiction, key management vulnerability and the like in traditional desensitization technology, so as to realize the balance between privacy protection and data usability.
Need to check novelty before this filing date? Find Prior Art