Decoupling Security Policies via Sensitive Types
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data security approaches in computer systems are inefficient due to the tightly coupled nature of security policies with target data, leading to proliferation of policies, lack of portability, reusability, and manageability, as well as increased administrative effort and risk of non-compliance.
Innovation Solution
Implementing a loosely coupled approach where security policies are indirectly associated with target data through sensitive types, allowing multiple data sets to be protected by a single policy, enabling decoupling from target data and facilitating portability, reusability, and efficient management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If security policies are tightly coupled with target data, then security protection can be directly applied to specific data, but the number of policies proliferates and management becomes complex
Solution Approach 1:
The patent introduces sensitive types as an intermediary layer between security policies and target data. Instead of directly coupling policies with specific data columns, policies are associated with sensitive types (e.g., SSN, credit card numbers), which then map to multiple data columns across different tables. This intermediary approach reduces the number of policies needed while maintaining effective security protection.
Solution Approach 2:
The patent makes security policies universal by associating them with sensitive types rather than specific data instances. A single security policy can be applied to multiple data columns that share the same sensitive type, allowing one policy to serve multiple functions across different tables and columns, thereby reducing policy proliferation.
2Reliability
If separate security policies are created for each security technology (encryption, access control, auditing), then each security requirement can be individually enforced, but the number of policies increases and portability decreases
Solution Approach 1:
The patent merges multiple security technologies (encryption, access control, auditing) into a unified security policy framework. Instead of creating separate policies for each security technology, all security requirements are consolidated into a single policy associated with a sensitive type, improving portability while maintaining individual enforcement capability.
Solution Approach 2:
The patent creates universal security policies that can enforce multiple security technologies simultaneously. A single policy can include encryption requirements, access control rules, and auditing configurations, making the policy portable and reusable across different data sets while maintaining comprehensive security enforcement.
3Reliability
If security policies are created for each data column requiring protection, then security can be applied to all sensitive data, but administrative effort increases and compliance management becomes difficult
Solution Approach 1:
The patent uses sensitive types as mediators to reduce administrative effort. Instead of creating and managing separate policies for each data column, administrators create policies for sensitive types, which automatically apply to all columns of that type across the database, significantly reducing the time and effort required for compliance management.
Solution Approach 2:
The patent enables preliminary classification of data into sensitive types, allowing security policies to be pre-configured and automatically applied. By categorizing data columns by their sensitive type in advance, the system automates policy application, reducing administrative effort and ensuring comprehensive protection without manual policy creation for each column.
4Reliability
If security policies are tightly coupled with target data, then security protection is directly applicable, but policies lack reusability and adaptability to changing requirements
Solution Approach 1:
The patent introduces sensitive types as intermediaries that decouple policies from specific target data. Policies associated with sensitive types can be reused across multiple data columns and tables, enhancing reusability and adaptability while maintaining direct applicability through the sensitive type mapping mechanism.
Data Source
AI summary
Techniques are provided for a data security system that includes two mappings: a first mapping that maps a security policy to sensitive type and a second mapping that maps the sensitive type to one or more data sets. The sensitive type indicates a class of sensitive data. Example data sets include columns, tables, tablespaces, files, and directories in a file system. Because a security policy is not tightly coupled to a target data set, the security policy becomes data-agnostic, portable, and reusable. Also, a security policy may be objectless in that, at some point in time, the security policy is not associated with any data set. A security policy may also be multifunctional in that the security policy may include multiple security features or requirements. A security policy may also be exhaustive in that all necessary security requirements prescribed for a data set can be included in the security policy.