Anonymized Data Comparison via Variant Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Comparing anonymized data entries is challenging due to variations in representation, making exact matching difficult, especially when names are encrypted, as fuzzy matching techniques are not feasible in such scenarios.
Innovation Solution
A system and method for comparing anonymized data entries involve normalizing, generating variants, scoring, anonymizing, and comparing these variants to enable exact matching between anonymized data entries, using a list expansion module, anonymizer module, and matching module to facilitate accurate comparison.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fuzzy matching techniques are used to compare names, then matching accuracy for variations is improved, but the method becomes inapplicable when names are anonymized or encrypted
Solution Approach 1:
The system performs preliminary actions by generating multiple variants of each anonymized name entry before comparison. These variants include different encodings, normalizations, and representations of the same anonymized data, so that when comparing against the anonymized database, at least one variant will match the database entry exactly, enabling fuzzy matching capability to be restored in the anonymized domain.
Solution Approach 2:
The system introduces an intermediary layer of variants that mediates between the anonymized query and the anonymized database. Instead of directly comparing the anonymized names (which may not match due to representation variations), the system generates intermediate variants that can bridge the gap and enable exact matching to occur.
2Reliability
If exact matching is used for anonymized data, then privacy protection is maintained, but matching accuracy decreases due to representation variations
Solution Approach 1:
The system performs preliminary variant generation on the anonymized query names before comparison. By expanding each anonymized name into multiple variants (different encodings, normalizations, representations), the system ensures that when these variants are compared against the anonymized database, exact matches can be found even though the original anonymized representations differed, thus maintaining both privacy and accuracy.
3Adaptability or versatility
If multiple variants are generated for each data entry, then matching capability for anonymized data is improved, but computational complexity increases
Solution Approach 1:
The system generates a limited but sufficient number of variants for each anonymized name entry rather than exhaustively generating all possible variants. This partial action approach creates enough variants to capture the essential matching capability while controlling computational resources. The variant generation focuses on the most relevant transformations (encodings, normalizations) rather than all theoretical possibilities.
Data Source
AI summary
Systems, methods and computer program products for the comparison of anonymized data entries are described. In one embodiment the present invention is a computer-implemented method for comparing anonymized data entries, including separately processing data entries in a query list and a search database by: normalizing, generating variants, scoring, and anonyminizing. The expanded and anonymized data entries in the query list and the search database are then compared to find an exact match. In another embodiment, the present invention is a system for comparing anonymized data entries, having a list expansion module, an anonymizer module, and a matching module.