Anonymization Framework for PII Data Privacy Protection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wireless telecommunication carriers face challenges in efficiently protecting personally identifiable information (PII) when sharing data with third parties, as existing methods fail to ensure effective privacy protection and result in uneven distribution of anonymization work, leading to potential data breaches.
Innovation Solution
A data privacy protection system that includes a landing zone for PII and non-PII data, listeners to store data in a database, agents to exclude PII from data sets, and an anonymization framework using hashing algorithms and sticky tables to create anonymized data sets, distributed across queues based on hashed device identifiers, ensuring efficient and secure data sharing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If PII data is shared with third parties, then data utility and analytics capabilities are improved, but privacy protection and security are worsened
Solution Approach 1:
The system extracts PII data from datasets before sharing with third parties. Listeners identify and separate PII fields (names, addresses, phone numbers, etc.) from non-PII data, creating distinct PII datasets that are then anonymized. This extraction ensures that only necessary non-identifying data is shared, protecting privacy while maintaining data utility for analytics.
Solution Approach 2:
The system introduces an intermediary anonymization framework between the data source and third-party recipients. This framework applies hashing algorithms to PII data, transforming it into anonymized form that cannot be traced back to individuals. The intermediary layer enables data sharing while enforcing privacy protection through cryptographic transformation.
2Reliability
If traditional anonymization methods are used, then privacy protection is improved, but workload distribution becomes uneven and processing efficiency is worsened
Solution Approach 1:
The system segments the anonymization workload by distributing it across multiple queues based on hashed device identifiers. Each queue handles a specific portion of the anonymization work, preventing any single processor from being overwhelmed. This segmentation enables parallel processing and balanced workload distribution, improving overall processing efficiency while maintaining comprehensive privacy protection.
Data Source
AI summary
A data privacy protection system is disclosed that comprises listener(s) that receive and store data including non-personal identifiable information (PII) and PII in data sets in a database and agent(s) that access each data set from the database, obtain the non-PII data and exclude the PII data to create non-PII data sets, and transmit the non-PII data sets to a third-party server. The system further comprises an anonymization framework that obtains the PII data from the data sets and stores some of the PII data in a raw PII data set. The anonymization framework distributes anonymization work on the stored PII data to queues based on hashed device identifiers associated with the stored PII data, performs the anonymization work on the stored PII data according to the queues to create an anonymized PII data set, and transmits the anonymized PII data set to the third-party server.