Encrypted Data Intersection for Privacy-Preserving Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in computing relationships between private data from different entities while preserving privacy and preventing data sharing, as entities are reluctant to share their data due to concerns about privacy and liability.
Innovation Solution
A method is developed where a server determines the intersection of member lists from two entities without accessing the lists directly, computes coefficients of a numeric relationship using information associated with the intersection, and provides a digital transmission representing these coefficients, ensuring that the private data remains secure and unshared.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If entities share their private data with each other to compute relationships, then the accuracy and value of data analysis is improved, but the privacy security and liability control of entities deteriorates
Solution Approach 1:
A trusted server acts as an intermediary that receives encrypted data from multiple entities, performs the intersection computation and relationship analysis on the encrypted data, and returns results without ever accessing the plaintext private data of any entity. This mediator enables collaborative analysis while preserving privacy security.
Solution Approach 2:
The patent transforms the data from plaintext to encrypted form using cryptographic transformations before processing. By changing the parameter state of the data (from readable to encrypted), the system enables computation while maintaining privacy, as the server operates on transformed parameters rather than original sensitive values.
2Object-affected harmful factors
If entities do not share their private data to preserve privacy, then privacy security is maintained, but the ability to compute relationships and gain insights deteriorates
Solution Approach 1:
The trusted server mediates the computation process by receiving encrypted inputs from entities, performing the intersection and relationship computation on encrypted data, and returning encrypted results. This allows relationship computation capability to function while entities maintain privacy security through encryption.
Solution Approach 2:
The patent replaces the mechanical process of data sharing and manual analysis with a cryptographic system. Instead of entities physically sharing data and manually computing relationships, the system uses encrypted computation protocols to automatically compute relationships on encrypted data, preserving privacy while enabling analysis.
3Measurement precision
If a server accesses member lists and private data directly to compute relationships, then the computation accuracy is improved, but the privacy protection and data security deteriorates
Solution Approach 1:
The server operates on encrypted data rather than plaintext data, changing the parameter state of the input. The cryptographic transformations allow the server to perform accurate computations on the encrypted representations of member lists and private data without ever accessing the actual sensitive values, maintaining both computation accuracy and data security.
Solution Approach 2:
The server functions as a trusted intermediary that processes encrypted data from entities. It performs the intersection computation and relationship analysis on the encrypted inputs, achieving accurate results while maintaining data security through the encryption layer that prevents direct access to sensitive information.
Data Source
AI summary
Aspects of the present disclosure relate to cryptography. In particular, example embodiments relate to computing a relationship between private data of a first entity and private data of a second entity, while preserving privacy of the entities and preventing inter-entity data sharing. A server includes a first component to compute an intersection of two datasets, without directly accessing either dataset. The server includes a second component to compute a relationship, such as a regression, between data in the first dataset and data in the second dataset, without directly accessing either dataset.