Privacy Data Integration via Generative Synthetic Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Companies face difficulties in integrating customer data across different systems while ensuring privacy protection, as existing methods may leak sensitive information during the integration process.
Innovation Solution
A privacy data integration method using generative models to create synthetic data from first and second privacy data, which are integrated using algorithms like database join, record linkage, or statistical match, ensuring privacy protection by maintaining similar joint probability distributions without directly transmitting the original data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If direct data integration methods are used to integrate customer data from different companies, then data integration capability is improved, but privacy security deteriorates due to potential leakage of sensitive information
Solution Approach 1:
The patent creates synthetic copies of original privacy data through generative models. These synthetic data copies preserve the statistical characteristics and joint probability distributions of the original data while containing no actual sensitive information. The synthetic data can be freely shared and integrated without privacy concerns, as it is merely a statistical replica rather than the real data itself.
Solution Approach 2:
The patent introduces synthetic data as an intermediary between the original privacy data and the data integration process. Instead of directly integrating sensitive original data, the system first transforms it into synthetic data through generative models, then performs integration operations on these intermediaries. This intermediary layer enables data sharing while maintaining privacy protection.
2Productivity
If traditional data integration algorithms are used to combine data from different sources, then integration efficiency is improved, but data quality and privacy protection deteriorate
Solution Approach 1:
The patent performs preliminary transformation of original privacy data into synthetic data before the actual integration process. By pre-processing the data through generative models to create privacy-preserving synthetic versions, the system enables subsequent efficient integration operations without compromising privacy. This preliminary action separates the privacy protection concern from the integration efficiency requirement.
3Measurement precision
If original privacy data is directly shared between companies for integration, then data accuracy is improved, but privacy security worsens due to exposure of sensitive information
Solution Approach 1:
The patent transforms the data representation parameters through generative models. The synthetic data preserves the statistical parameters and joint probability distributions that define data accuracy, while changing the actual data values to synthetic representations. This parameter transformation maintains the analytical value and accuracy characteristics needed for integration while eliminating the exposure of actual sensitive information.
Data Source
AI summary
A privacy data integration method and a server are provided. The privacy data integration method includes the following steps. A first processing device and a second processing device respectively obtain a first generative model and a second generative model according to a first privacy data and a second privacy data. A server generates a first generative data and a second generative data via the first generative model and the second generative model respectively. The server integrates the first generative data and the second generative data to obtain a synthetic data.