Anonymization Server for Click Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Click fraud, which involves artificial clicks on web advertisements to inflate traffic statistics or trigger revenue recognition, is challenging to detect without compromising the privacy of legitimate users, as existing methods rely on IP address tracking that also profiles legitimate users.
Innovation Solution
A method that generates an anonymous representation of a client device's unique identifier using hashing operations, incorporating a time period, to anonymize requests and detect fraudulent clicks without revealing the original identifier, allowing for secure tracking and fulfillment of legitimate requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If IP address tracking is used to detect click fraud, then click fraud detection capability is improved, but user privacy is compromised due to profiling of legitimate users
Solution Approach 1:
The patent introduces an intermediary anonymization server that sits between the client device and the ad serving system. This server receives requests with identifying information, strips or transforms the identifying data into anonymous identifiers, and forwards the anonymized requests. The intermediary enables fraud detection through pattern recognition on anonymous data while preventing direct profiling of individual users, thus resolving the contradiction between detection capability and privacy protection.
Solution Approach 2:
The system creates anonymous copies of user requests by generating pseudonymous identifiers that replicate the functional need for tracking without containing actual personal identifying information. These anonymous identifiers serve as substitutes for real user identities, allowing the system to track click patterns and detect fraud while maintaining user anonymity. The copying principle enables the decoupling of detection functionality from privacy-invasive tracking.
2Loss of information
If anonymous identifiers are generated using hashing operations, then user privacy is preserved, but the ability to track legitimate user behavior for fraud detection may be reduced
Solution Approach 1:
The system performs preliminary anonymization of user identifiers before requests are processed by the ad serving system. By pre-converting identifying information into anonymous identifiers through hashing operations, the system establishes a consistent anonymous identity for each user session. This preliminary action enables subsequent fraud detection algorithms to track behavior patterns across multiple requests using the anonymous identifiers, maintaining detection reliability without compromising privacy.
Solution Approach 2:
The patent transforms the identifying parameters from real user identifiers to anonymous identifiers through hashing operations. This parameter change maintains the functional utility of identification for tracking purposes while fundamentally altering the nature of the identifier to remove personally identifiable information. The transformed parameters enable fraud detection through pattern recognition while preserving user anonymity, resolving the contradiction between privacy preservation and detection accuracy.
Data Source
AI summary
A click detection method, apparatus and system is provided. An embodiment includes a method that receives a request from a client destined for a server and a unique identifier for the client. The method also includes generating a representation of the unique identifier, and forwarding the representation of the unique identifier and the request to the server.