Encrypted Data Analytics Platform Decoupling Privacy and Computation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing security protocols for data analytics are inflexible and unsuitable for a wide range of analytic tasks, often intertwining privacy protection with data analytics, making it difficult to perform diverse tasks while maintaining dataset privacy.
Innovation Solution
A unified platform decouples privacy protection from data analytics, using encryption protocols to map private datasets from multiple vendors into a common space where diverse analytic tasks can be performed by a mediator without exposing the data, allowing for secure and flexible data analytics across various functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing security protocols are used for data analytics, then data privacy is protected, but the protocols are inflexible and unsuitable for a wide range of analytic tasks
Solution Approach 1:
The patent segments the data analytics process into distinct components: encrypted data submission, encrypted computation execution, and result retrieval. This allows different analytic operations to be performed on encrypted data without requiring decryption, enabling versatility while maintaining privacy protection through modular cryptographic operations
Solution Approach 2:
The patent introduces an intermediary cryptographic processing layer that mediates between the data owner and the analytics system. This intermediary enables computations on encrypted data by translating analytic tasks into cryptographic operations, resolving the contradiction between privacy protection and analytic flexibility
2Reliability
If privacy protection is intertwined with data analytics, then data security is maintained, but it becomes difficult to perform diverse analytic tasks
Solution Approach 1:
The patent separates privacy protection mechanisms from data analytics operations by implementing encryption/decryption as distinct modular components. This segmentation allows diverse analytic tasks to be performed on encrypted data without being constrained by privacy protection requirements, as the cryptographic layer independently handles security while the analytics layer focuses on computational flexibility
3Adaptability or versatility
If a unified platform decouples privacy protection from data analytics, then diverse analytic tasks can be performed, but system complexity increases
Solution Approach 1:
The patent implements a universal cryptographic processing framework that handles multiple analytic operations (aggregation, filtering, sorting, etc.) through a common encrypted computation mechanism. This multi-functionality reduces system complexity by consolidating diverse analytic capabilities into a single unified platform that processes all tasks through the same cryptographic infrastructure
Data Source
AI summary
Data analytics on encrypted data elements is disclosed. One example is a system including a first data system, a second data system, and a data analytics system. The first data system includes a first data element and a first encryption module with a first private key. The second data system includes a plurality of second data elements and a second encryption module with a second private key. The first encryption module and the second encryption module are communicatively linked to one another, to apply, via the first and second private keys, an encryption protocol to the first data element and the plurality of second data elements to encrypt the data elements. The data analytics system maps the encrypted data elements to an analytics space, performs data analytics based on the mapped data elements, and distributes, via a computing device, results of the data analytics to an information retrieval system.