Data Clean Room Query Validation With Differential Privacy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in securely and efficiently sharing database data between different users, particularly in creating target groups for advertising or marketing efforts, due to privacy concerns and the difficulty in performing data analysis like overlap analysis across disparate datasets.
Innovation Solution
A clean room system implements query templates with differential privacy mechanisms to obfuscate shared data, using epsilon budgeted noise injection, and row access policies to manage secure data sharing, allowing multiple parties to analyze data while protecting privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If database data is shared between different users, then data analysis capability is improved, but security and privacy protection deteriorate
Solution Approach 1:
The patent introduces a data clean room as an intermediary environment that enables secure data sharing between users. The clean room acts as a mediator where data can be analyzed without exposing raw sensitive information, thus improving data analysis capability while maintaining security and privacy protection through controlled access mechanisms.
Solution Approach 2:
The patent segments data access by implementing row access policies that divide data into accessible and protected portions. By segmenting the data sharing mechanism into different access levels and using query templates with parameterized access controls, the system enables analytical capabilities while preventing unauthorized exposure of sensitive information.
2Reliability
If query validation is performed on shared data, then data security is improved, but query processing time increases
Solution Approach 1:
The patent implements query template validation as a preliminary action before actual data querying. By pre-defining and validating query templates with proper access controls and security checks, the system establishes security rules in advance, allowing subsequent queries to execute faster without repeated validation overhead while maintaining data security.
3Reliability
If differential privacy mechanisms are applied to obfuscate shared data, then privacy protection is improved, but data analysis precision deteriorates
Solution Approach 1:
The patent applies differential privacy mechanisms selectively to specific columns and rows based on sensitivity requirements. By implementing row access policies that apply privacy obfuscation only where necessary while leaving other data accessible, the system achieves privacy protection for sensitive information while maintaining analysis precision for non-sensitive data.
Data Source
AI summary
Embodiments of the present disclosure may provide a data sharing system implemented as a local application of a distributed database. A query from a query template can be validated and executed against shared dataset that comprises portions of data from the database dataset and additional portions of data from another database of the distributed database.