Tobacco enterprise brain system data security protection method

By using descriptive metadata and blood-dependent multi-dimensional feature classification in the brain system of tobacco enterprises, combining the correlation between the multi-grained data access baseline model and the blood-related knowledge graph, the problem of difficulty in timely discovering and tracking of abnormal access behavior in the existing technology is solved, and more accurate abnormal access detection and traceability of cross-domain data access is achieved.

CN119557877APending Publication Date: 2025-03-04CHINA TOBACCO ZHEJIANG IND CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411741694.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-29
Publication Date
2025-03-04

AI Technical Summary

Technical Problem

Existing enterprise brain system data security protection methods are difficult to fully consider the semantic associations of massive heterogeneous data, making it difficult to detect and track abnormal access behaviors in a timely manner.

Method used

By determining the security domain to which the inflow data belongs based on descriptive metadata and blood relationship dependence, the correlation between the baseline model and blood relationship knowledge graph is used through multi-grained data access, the outliers of data access events are calculated, and the accuracy of abnormal access detection is improved.

Benefits of technology

It realizes differentiated configuration of security policies on different granularities, improves the accuracy of abnormal access time detection, and supports traceability of cross-domain data access.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119557877A_ABST
    Figure CN119557877A_ABST
Patent Text Reader

Abstract

The invention discloses a tobacco enterprise brain system data security protection method. The method comprises the following steps: acquiring inflow data and a data access event of a tobacco enterprise brain system; according to the inflow data, feature classification and edge weight calculation are carried out through descriptive metadata extraction and a data consanguinity analysis method, a security domain belonging to a pre-constructed consanguinity knowledge graph is obtained, and a corresponding data protection strategy is inherited; and based on a pre-constructed multi-granularity data access baseline model, calculating an abnormal value corresponding to the data access event, and if the abnormal value is greater than a preset abnormal threshold, determining that the data access event is an abnormal access event and performing dynamic risk assessment. According to the method, the security domain to which the inflow data belongs is determined on the basis of descriptive metadata and blood dependency data multi-dimensional feature classification, so that the security policy can be subjected to differential configuration on different granularities. In addition, a multi-granularity data access baseline model is used, so that the accuracy of abnormal access time detection is improved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Sensitive data security compliance processing system and method

    CN120316821A

  • AI data warehouse full-link consanguinity tracking method and AI data warehouse full-link consanguinity tracking device

    CN120353843A