PII Anonymization via Poly-Identifiers for Secure Data Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data storage systems face challenges in maintaining confidentiality, particularly with medical information, as HIPAA regulations restrict the sharing and collation of personal identifiable information (PII), limiting cross-indexing and correlation analysis across datasets.
Innovation Solution
A method involving parsing PII into separate fields and replacing it with anonymous identifiers, both internally and externally keyed, to create a correlation score for matching across different data sources while ensuring data security and compliance with HIPAA regulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If PII is stored in structured data stores for easy access and analysis, then data availability and analytical capability are improved, but data confidentiality and privacy protection deteriorate
Solution Approach 1:
The patent introduces an intermediary system that sits between the data storage and access points, translating queries using poly-identifiers without exposing underlying PII. This mediator layer enables data analysis while maintaining confidentiality by preventing direct access to raw personal information.
Solution Approach 2:
The patent extracts PII from the data structure and replaces it with poly-identifiers, separating the identification function from the personal information. This extraction allows data to be analyzed and correlated without the harmful element of exposed PII.
2Adaptability or versatility
If PII is shared and collated across organizations for beneficial data access, then data utility and research capability are improved, but compliance with HIPAA privacy requirements deteriorates
Solution Approach 1:
The patent creates a universal poly-identifier system that can be applied across multiple organizations and data types while maintaining HIPAA compliance. This universal approach enables data sharing across different entities without requiring organization-specific solutions, thereby improving adaptability while ensuring regulatory reliability.
3Object-affected harmful factors
If PII is scrubbed from datasets before sharing to maintain privacy, then data confidentiality is improved, but the ability to cross-index and correlate datasets deteriorates
Solution Approach 1:
The patent creates copies of identification functionality through poly-identifiers that preserve the ability to correlate data without copying actual PII. These identifier copies maintain the relational structure needed for cross-indexing while replacing sensitive information with anonymous but linkable tokens.
4Object-affected harmful factors
If anonymous identifiers are used to replace PII for security, then data confidentiality is improved, but the ability to match records across sources deteriorates
Solution Approach 1:
The patent implements dynamic poly-identifiers that can adapt their matching criteria based on the context and data being compared. This dynamic approach allows the system to maintain high matching accuracy by adjusting the stringency of identifier comparison according to the specific data domain and correlation needs, while still preserving security through anonymous identification.
Data Source
AI summary
A method for data security including receiving a first recordset, said first recordset including a first poly-identifier representing a first personally identifiable information (PII), and a first contextual information, said first poly-identifier associated with a name field of a record in a PII structured data store. Also receiving at the server a second recordset, said second recordset including a second poly-identifier representing a second personally identifiable information (PII) and a second contextual information, said second poly-identifier comprised of unique characters associated with the name field of a record in the PII structured data store. Then comparing the first and second contextual information to calculate a correlation score to create a match table entry as a result of said comparing, said match table entry including both an internal ID and an external anonymous ID. The IDs may associate the contextual information between records to a single person.