Sensitive Data Image Encoding for Cross-Border Deduplication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data analytics systems require exposing sensitive data, such as PII, across jurisdictional boundaries, violating regulations and risking security breaches.
Innovation Solution
A system that encodes sensitive data into unique image representations using color mappings, allowing clustering and deduplication without exposing the original data, using unsupervised learning to identify similar data instances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensitive data is accessed and reviewed for data analytics operations, then data analytics accuracy is improved, but data security and regulatory compliance deteriorate
Solution Approach 1:
The patent creates image representations that are visual copies of sensitive data instances. These image representations contain the same information as the original sensitive data but in a transformed visual format. The image representations can be processed for analytics without exposing the actual sensitive data content, thus maintaining security while enabling accurate analysis.
Solution Approach 2:
The patent transforms sensitive data from its original format into image representations by changing the parameter representation format. Each sensitive data instance is converted into an image with specific visual parameters (colors, patterns, layouts) that encode the data information. This parameter transformation allows analytics operations to be performed on the visual representation without accessing the underlying sensitive data.
2Adaptability or versatility
If sensitive data is shared across jurisdictional boundaries for analytics, then data analytics capability is improved, but regulatory compliance deteriorates
Solution Approach 1:
The system creates image representations that serve as copies of sensitive data instances. These image copies can be transmitted and processed across different jurisdictions without violating data sharing regulations, as the actual sensitive data remains localized. The image representations contain sufficient information for analytics operations while not constituting prohibited data transfers.
Solution Approach 2:
The image representation acts as an intermediary between the sensitive data source and the analytics processing system. Instead of directly sharing sensitive data across jurisdictions, the system transforms data into image representations that serve as a mediator, enabling analytics capability while maintaining regulatory compliance by preventing direct exposure of sensitive data content.
3Object-affected harmful factors
If image representations with color mappings are used to encode sensitive data, then data security is improved, but system complexity increases
Solution Approach 1:
The patent applies parameter changes by transforming sensitive data into image representations with specific visual parameters. Each sensitive data instance is encoded using color mappings where different colors represent different data characteristics. This parameter transformation approach simplifies the security mechanism compared to traditional encryption methods, as it relies on visual encoding rather than complex cryptographic algorithms.
Solution Approach 2:
The system uses color changes as a core encoding mechanism. Sensitive data instances are represented by images with specific color schemes where colors encode data information. This color-based encoding provides security by making the data visually distinctive and non-trivial to reverse-engineer, while the implementation remains relatively simple compared to cryptographic approaches.
4Measurement precision
If manual review of sensitive data content is performed, then data analytics accuracy is improved, but time consumption and resource usage increase
Solution Approach 1:
The system creates image representations that capture the essential information of sensitive data instances in visual form. These image copies can be processed automatically by computer vision algorithms rather than requiring manual review. The image representations preserve the information needed for accurate analytics while enabling automated processing, thus reducing time consumption significantly.
Solution Approach 2:
The patent replaces the mechanical process of manual data review with automated image processing. Instead of human analysts manually examining sensitive data content, the system transforms data into image representations that can be processed by automated computer vision and machine learning algorithms. This substitution dramatically reduces time consumption and resource usage while maintaining or improving analytics accuracy.
Data Source
AI summary
Systems and methods for performing data analytics on sensitive data, such as deduplication are disclosed herein. A first request to perform a first data analytics operation for a set of sensitive data instances may be received. A first set of image representations for the set of sensitive data instances may be retrieved. A first sensitive data instance and a second sensitive data instance may be clustered into a first cluster. The first data analytics operation may be performed on the first cluster.