Privacy Budget Allocation for Multi-Attribute Data Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to appropriately determine the allocation of privacy budgets for multiple data aggregations, considering the degree of influence of noise on aggregation results, which is crucial for protecting privacy in statistical data utilization among multiple companies.
Innovation Solution
An information processing device that includes a basic aggregation information acquisition unit, an estimation unit, an evaluation unit, and an allocation determination unit to quantify the influence of noise for each combination in an aggregation pattern and optimize the privacy budget allocation based on these evaluations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If noise is added to aggregation results to protect privacy, then privacy protection is improved, but the accuracy of aggregation results deteriorates
Solution Approach 1:
The patent changes the parameter of privacy budget allocation from uniform to optimized distribution. By evaluating the influence degree of noise for different aggregation combinations and allocating privacy budgets accordingly, the system achieves better balance between privacy protection and result accuracy. The privacy budget is dynamically adjusted based on the specific characteristics of each aggregation operation.
2Productivity
If multiple aggregations are performed on the same dataset, then data utilization is improved, but the difficulty of determining privacy budget allocation increases
Solution Approach 1:
The patent performs preliminary evaluation of the influence degree of noise for each aggregation combination before actually performing the aggregations. This preliminary action includes calculating the sample size of aggregation results and evaluating how much noise will affect each combination. Based on these pre-evaluations, the system determines the optimal privacy budget allocation for multiple aggregations, avoiding the complexity of post-hoc adjustments.
Solution Approach 2:
The patent implements a feedback mechanism where the evaluation results of noise influence are used to adjust and optimize privacy budget allocation. The system continuously monitors the impact of noise on different aggregation combinations and refines the privacy budget distribution to achieve optimal balance between privacy protection and data utilization efficiency.
3Ease of operation
If uniform privacy budget is allocated to all aggregations, then simplicity is improved, but the optimality of privacy protection deteriorates
Solution Approach 1:
The patent applies local quality by allocating different privacy budget levels to different aggregation combinations based on their specific characteristics. Instead of uniform allocation, the system evaluates each aggregation combination's sensitivity to noise and assigns appropriate privacy budget levels. This localized approach ensures that aggregations with higher noise sensitivity receive more privacy budget, while less sensitive aggregations receive fewer resources, optimizing overall privacy protection.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing device (10A) includes: a basic aggregation information acquisition unit (14A) that acquires basic aggregation information of devices (10A, 10B), each holding user data including a user ID and attribute information related to a user; an estimation unit (15) that estimates sample size of an aggregation result based on the basic aggregation information and a basic value for estimating the sample size of the aggregation result; an evaluation unit (16) that evaluates influence of noise for each combination included in an aggregation pattern containing multiple combinations of data items included in the attribute information, based on the sample size; and an allocation determination unit (17) that determines an allocation of a privacy budget for noise for each combination for aggregation targeting the aggregation pattern, within a predetermined range of the privacy budget, based on evaluation results for each combination.