Anonymization Module Noise Pool Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for anonymizing database queries are inefficient in computation and memory usage, and fail to effectively protect individual user privacy against condition/inverse condition pairs, particularly when repeated queries are allowed.
Innovation Solution
A system with an anonymization module that adds noise elements to mask attack components in queries, using fixed-random numbers to ensure that information about individual users is highly unlikely to be inferred from single or repeated answers, and ignores database changes outside specified time epochs to maintain privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If noise elements are added to mask attack components in queries, then individual user privacy is protected, but computation efficiency and memory usage deteriorate
Solution Approach 1:
The system pre-generates noise elements and stores them in memory before queries arrive. By preparing the noise pool in advance rather than generating noise on-demand for each query, the system achieves differential privacy protection while maintaining efficient query processing speed.
Solution Approach 2:
The noise pool is divided into multiple segments or buckets, and different noise elements are drawn from different segments based on query characteristics. This segmentation allows efficient memory access patterns and enables the system to handle multiple queries without reprocessing the entire noise pool each time.
2Reliability
If noise elements are added to mask attack components in queries, then individual user privacy is protected, but memory consumption increases
Solution Approach 1:
The system pre-generates and stores noise elements in memory before they are needed. By preparing the noise pool in advance, the system can use efficient memory allocation and avoid the overhead of random noise generation during query processing, thereby controlling memory usage while maintaining privacy protection.
Solution Approach 2:
After noise elements are used to answer queries, they can be discarded or recycled. The system can replenish the noise pool by generating new noise elements only when the existing pool is exhausted, rather than maintaining a continuously growing memory allocation, thus optimizing memory consumption over time.
3Measurement precision
If the system allows repeated queries to maintain accuracy, then statistical analysis utility is improved, but privacy protection against condition/inverse condition pairs deteriorates
Solution Approach 1:
The system prepares multiple noise elements in advance that can be sequentially applied to the same query. When a user repeats a query, the system applies different pre-generated noise elements from the pool, ensuring that each repeated query receives a different noisy answer. This prevents attackers from eliminating noise through repeated queries while maintaining statistical accuracy through the use of calibrated noise parameters.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An anonymizing method for a database system is provided. The method includes receiving a query, identifying one or more attack components in the query, computing one or more noise elements from the one or more attack components, and computing fixed noise or fixed thresholds from the one or more noise elements.