Client Event PII Removal via Proxy Anonymization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Personal identifiable information (PII) is often inadvertently collected by applications without user consent, posing risks of identity theft and violating user agreements, as current technologies lack effective methods to securely remove PII from event data while maintaining user privacy.
Innovation Solution
An system that identifies and removes PII from client event information at the client or proxy level, transmitting anonymized event data to a server, using techniques such as overwriting or hashing personal data and partially redacting IP addresses to prevent identification, allowing for statistical analysis without revealing sensitive information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If applications collect event information for analysis, then useful statistical data is obtained, but personal identifiable information is inadvertently captured creating privacy risks
Solution Approach 1:
The patent extracts and removes personal identifiable information (PII) from event data at the source (client or proxy level) before transmission to the server. This extraction process separates useful event information from sensitive PII, allowing statistical analysis to proceed on anonymized data that retains analytical value while eliminating privacy risks.
Solution Approach 2:
The patent applies preliminary anonymization actions by removing PII before data leaves the client device or proxy server. Techniques such as hashing, overwriting, and partial redaction are performed in advance, ensuring that PII never enters the transmission pipeline to the central server, thus preventing privacy breaches before they can occur.
2Object-affected harmful factors
If PII is removed from event data, then user privacy is protected, but data completeness and analytical value may be reduced
Solution Approach 1:
The patent applies different quality treatments to different parts of the event data structure. Non-PII fields retain their original quality and completeness for full analytical value, while PII fields receive targeted anonymization treatment. This local differentiation preserves data completeness for analysis while protecting privacy in sensitive areas.
Solution Approach 2:
The patent transforms PII parameters through hashing functions and redaction techniques that change the data format while preserving the ability to perform statistical analysis. For example, IP addresses are partially redacted to remove location-specific precision while maintaining network-level aggregation capability, thus changing parameters to balance privacy and analytical utility.
3Object-affected harmful factors
If PII removal techniques are implemented at client level, then privacy is enhanced, but device complexity and processing overhead increase
Solution Approach 1:
The patent introduces a proxy server as an intermediary that can perform PII removal centrally, reducing the complexity burden on individual client devices. The proxy acts as a mediator between the client and the central server, handling the computationally intensive anonymization tasks while clients only need to transmit raw event data, thus distributing complexity away from resource-constrained devices.
4Object-affected harmful factors
If all PII is completely removed from event data, then maximum privacy protection is achieved, but ability to trace and correlate events from the same user is lost
Solution Approach 1:
The patent applies partial redaction rather than complete removal of all identifying information. For example, IP addresses are partially redacted to remove the last octet, which provides sufficient privacy protection for individual identification while retaining enough information to correlate events at the network level. This partial action achieves the optimal balance between privacy and correlation capability.
Data Source
AI summary
Event information with respect to events that occur at a client is identified. At a client and/or a proxy, personal identifiable information (PII) from the event information is removed to produce event data. The event data is transmitted to a server.