Data Security Grading via Lineage Tree Tracing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional rule-based methods for data security grading are inadequate in handling the rapid expansion of data sizes and changes in dataset content, leading to inefficient and potentially risky security level assignments.
Innovation Solution
The implementation of a system that uses a lineage tree to trace derived data fields to their source data fields, allowing for the automatic propagation and dynamic adjustment of security levels based on derivation relations, eliminating the need for constant rule updates and reducing reliance on table or field names.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional rule-based methods are used for data security grading, then security levels can be assigned to data fields, but the methods become inadequate and inefficient when handling rapid expansion of data sizes and changes in dataset content
Solution Approach 1:
The system pre-computes and stores security levels for source data fields in a security level table before queries arrive. When data fields are created or updated, their security levels are determined in advance and cached, eliminating the need to re-evaluate security rules for every query. This preliminary action enables rapid security level retrieval even as data volumes expand.
Solution Approach 2:
The patent implements a nested structure where derived data fields contain references to parent data fields, which in turn contain references to their parents, ultimately tracing back to source data fields. This nested lineage tree structure allows the system to inherit security levels from source fields through multiple levels of derivation, providing adaptability to complex data relationships without requiring explicit rules for each derived field.
2Reliability
If rule-based methods are used to handle changing dataset content, then security rules can be defined, but constant re-writing of rules is required when data content changes
Solution Approach 1:
The system implements feedback through the lineage tree structure that automatically traces derived data fields back to their source fields. When source data fields change or are updated, the system can automatically propagate security level changes through the lineage tree to all derived fields, eliminating manual rule updates. The security level table is dynamically updated based on changes in source fields, providing continuous feedback loops that maintain accuracy without requiring constant human intervention.
3Measurement precision
If security levels are manually assigned to each data field, then proper security classification can be achieved, but the process becomes extremely challenging with large-scale online operations generating new data constantly
Solution Approach 1:
The system implements self-service by automatically determining security levels for derived data fields through the lineage tree structure. When a query accesses a derived data field, the system automatically traces the lineage to source fields, retrieves their security levels from the security level table, and determines the appropriate security level for the derived field without human intervention. This self-service mechanism maintains precise security classification across thousands of data fields while reducing operational complexity.
Data Source
AI summary
Systems and methods are provided for data security grading. An exemplary method for data security grading, implementable by a computer, may comprise receiving a request to access a query data field, searching for the query data field from a security level table, in response to finding the query data field from the security level table, obtaining from the security level table a security level corresponding to the query data field, and in response to not finding the query data field from the security level table, determining a security level corresponding to the query data field based at least on a lineage tree and the security level table. The lineage tree may trace the query data field to one or more source data fields, and the security data level table may comprise one or more security levels corresponding to the one or more source data fields.