In-Database Redaction for Semi-Structured Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional mechanisms for redacting sensitive information in databases often require exporting data, which can expose sensitive information and create security and governance issues. Additionally, these methods are limited to structured data and do not effectively handle semi-structured and unstructured data.
Innovation Solution
The system performs data mapping to identify sensitive data within databases, allowing for in-situ redaction without exporting the data. This includes generating path mappings for semi-structured data and index mappings for unstructured data, enabling selective masking of sensitive information at a sub-column level.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional redaction mechanisms export data to redact sensitive information, then redaction can be performed, but security risks increase and governance complexities arise
Solution Approach 1:
The patent introduces an intermediary redaction layer between the database and users. Instead of exporting sensitive data for redaction, the system implements in-database redaction mechanisms that mask sensitive information while allowing legitimate access to non-sensitive data. This intermediary approach prevents sensitive data from leaving the secure database environment while still enabling redaction functionality.
Solution Approach 2:
The patent creates a copy of the database data with redactions already applied. Rather than exporting original sensitive data for external redaction, the system generates a redacted copy that can be safely accessed by users. This copying mechanism ensures the original sensitive data remains protected in the source database while providing a safe alternative for user access.
2Adaptability or versatility
If conventional rediction methods are used, then structured data can be redacted, but semi-structured and unstructured data cannot be effectively handled
Solution Approach 1:
The patent implements a universal redaction system that handles multiple data types (structured, semi-structured, and unstructured) through a single integrated mechanism. The system uses data mapping and pattern recognition capabilities that work across different data formats, making the redaction functionality applicable to diverse data structures without requiring separate specialized processes for each type.
Solution Approach 2:
The patent changes the parameters of the redaction system to accommodate different data types. By implementing flexible data mapping approaches and adjustable redaction rules, the system can adapt its behavior based on the structure and characteristics of the input data, whether structured, semi-structured, or unstructured, thereby maintaining effective redaction across all data types.
3Reliability
If data is exported for redaction, then sensitive information can be masked, but data governance complexities increase
Solution Approach 1:
The patent implements self-service redaction capabilities within the database system itself. The database automatically identifies and redacts sensitive information using built-in data mapping and recognition mechanisms, eliminating the need for external export and manual redaction processes. This self-service approach reduces governance complexity by consolidating rediction functions within the existing database infrastructure.
Solution Approach 2:
The patent performs preliminary data mapping and sensitivity identification before data access occurs. By pre-configuring redaction rules and mapping sensitive data elements, the system prepares the redaction framework in advance, reducing the complexity of ongoing governance. This preliminary action ensures that rediction is automatically applied according to predetermined policies without requiring complex real-time governance decisions.
Data Source
AI summary
The present disclosure describes systems, methods, and computer program products for redacting sensitive data within a database. An example method can include sending, to a database, a data query accessing a column of the database, a masking policy identifying a first category of sensitive data and a second category of sensitive data, and in response to the data query, receiving redacted data, wherein the first category of sensitive data is redacted from a first location of the column by a first redaction operation and the second category of sensitive data is redacted from a second location of the column by a second redaction operation.