Lightweight Entity Database for Sensitive Data Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Organizations face challenges in protecting sensitive structured data from unauthorized disclosure due to the limitations of existing methods, such as Wu-Manber and Aho-Corasick, which do not scale well for large datasets, and full-blown databases being unsuitable for laptops, leading to security concerns and false positives when using pattern inference and regular expressions.
Innovation Solution
Implementing protect agents across egress points to identify and secure sensitive entities by converting them into canonical formats and storing them in lightweight entity databases (LWED) or global entity databases (GED), which support compressed versions and membership queries, allowing for secure comparison and initiation of security actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If full-blown databases are employed for exact data matches, then search capability is improved, but device complexity and security concerns increase, making them unsuitable for laptops
Solution Approach 1:
The patent extracts only the essential functionality needed for exact data match detection by implementing a lightweight database structure that stores only critical data elements (such as social security numbers, patient IDs) rather than complete database copies. This extraction approach maintains detection capability while dramatically reducing complexity for mobile devices like laptops.
Solution Approach 2:
The invention uses a simplified, lightweight database structure that can be easily implemented on resource-constrained devices. Instead of maintaining complex full-blown databases, the system employs a streamlined data storage mechanism that provides sufficient functionality for security purposes while minimizing overhead and complexity.
2Adaptability or versatility
If pattern inference and regular expressions are used for data matching, then adaptability is improved, but false positives increase and security is compromised
Solution Approach 1:
The patent replaces pattern inference and regular expression matching with a direct exact match comparison approach. Instead of using complex pattern recognition algorithms that are prone to false positives, the system stores canonical representations of sensitive data elements and performs precise equality comparisons, thereby eliminating false positives while maintaining adaptability through configurable data element identification.
3Speed
If Wu-Manber and Aho-Corasick methods are used for keyword matching, then search speed is improved, but scalability deteriorates for large datasets
Solution Approach 1:
The invention changes the fundamental parameter of data storage from storing complete documents or large datasets to storing only extracted data elements (such as individual social security numbers, patient IDs, or other sensitive information tokens). This parameter change enables the system to maintain fast search speeds through exact match comparisons while achieving scalability, as the lightweight database contains only the essential data elements rather than entire large datasets.
Data Source
AI summary
A technique for preventing selected sets of data words from unauthorized transmission out of the secure perimeter of a computer system is disclosed. A set of security rules is applied to an outgoing data message and if one of the set of rules is triggered, scanning by another set of security rules. The server then executes the security command before transmitting the outgoing message out of the secure perimeter of the computer system or blocking transmission.