Encrypted Personal Data Matching with Double-Partition Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for comparing biometric data in an encrypted domain, such as in Fully Homomorphic Encryption (FHE), are computationally burdensome due to the high cost of bootstrapped operations like comparisons, making real-time identification and database queries impractical.
Innovation Solution
A method involving a double partition of a reference personal database into first and second sets, followed by applying a strictly convex function to similarity rates and calculating overall similarity rates for each set, then comparing these rates with predetermined thresholds, significantly reducing the number of bootstrapped operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If homomorphic encryption is used to perform comparisons in the encrypted domain, then user privacy and database confidentiality are guaranteed, but computational burden increases significantly making real-time identification impractical
Solution Approach 1:
The reference database is partitioned into multiple subsets (first partition into first sets, second partition into second sets). Each subset is processed independently to compute partial similarity rates, which are then aggregated. This segmentation reduces the number of bootstrapped operations from O(m*n) to O(m+n), enabling real-time processing while maintaining encrypted domain security.
Solution Approach 2:
Instead of computing exact similarity rates for all reference data entries, the method computes approximate overall similarity rates by aggregating partial results from subsets. This partial action approach provides sufficient accuracy for identification while dramatically reducing computational complexity and bootstrapped operation count.
2Measurement precision
If all reference data are processed individually in the encrypted domain, then identification accuracy is maintained, but the number of bootstrapped operations becomes prohibitively large
Solution Approach 1:
Similarity rates from multiple reference data entries within each subset are merged by computing an overall similarity rate through aggregation (summarization). This combining approach maintains identification accuracy by preserving the collective information from all references while reducing the number of individual comparison operations required.
Solution Approach 2:
The problem is transformed from comparing candidate data with each reference entry individually (one-dimensional approach) to comparing aggregated subset results (two-dimensional approach with first and second partitions). This dimensional change reduces computational complexity from quadratic to linear relationship.
Data Source
AI summary
A method for processing personal data, comprising the steps of: (a) For each reference personal data of a reference personal database, calculating in the encrypted domain a similarity rate of the reference personal data with a candidate personal data; said reference personal database being associated with a first partition into a plurality of first sets of reference personal data, and with a second partition into a plurality of second sets of reference personal data, such that each reference personal data of a reference personal database belongs to a single first set and a single second set; (b) For each first set and each second set, calculating an overall similarity rate of said set based on the similarity rates of the reference personal data of said set; (c) Comparing each overall similarity rate of a first and second set with a first and second predetermined threshold, respectively.