Privacy-Preserving Cookie Generator Using Bloom Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online services face a challenge in providing personalized experiences while maintaining user privacy, as traditional tracking methods compromise privacy and are inefficient, and existing obfuscation techniques like noise addition or generalization come with high communication overhead and reliance on noise dictionaries.
Innovation Solution
The implementation of a privacy-preserving cookie generator that encodes user profiles into a Bloom filter with naturally occurring noise, allowing for personalized services without revealing sensitive information, by injecting noise into the encoded profile to create a privacy-preserving data structure that can be sent over a network, enabling personalized results while maintaining user privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional tracking methods (cookies, device fingerprinting) are used to provide personalized services, then personalization quality is improved, but user privacy is compromised
Solution Approach 1:
A trusted third party (TTP) is introduced as an intermediary to generate privacy-preserving cookies that contain obfuscated user profile information. The TTP creates noisy versions of user profiles that preserve personalization capability while preventing direct tracking, acting as a mediator between the user's privacy concerns and the service provider's personalization needs
Solution Approach 2:
The user profile data is transformed by adding controlled noise and obfuscation parameters. The privacy-preserving cookie contains a noisy version of the user profile where sensitive information is obscured through parameter modification, allowing personalization while maintaining privacy
2Object-affected harmful factors
If noise addition techniques are used to obfuscate user profiles, then user privacy is improved, but communication overhead increases due to reliance on noise dictionaries
Solution Approach 1:
The requirement for external noise dictionaries is extracted and eliminated from the system. The privacy-preserving cookie generation process uses only the user's actual profile information and randomly selected items from the profile itself, removing the need for separate noise dictionary resources and reducing communication overhead
Solution Approach 2:
The user profile itself serves as the source of obfuscation material. Random items are selected from the user's own profile to create the noisy version, making the system self-sufficient and eliminating dependency on external noise dictionaries
3Object-affected harmful factors
If generalization techniques are used to obfuscate user profiles, then user privacy is improved, but personalization accuracy deteriorates
Solution Approach 1:
Instead of generalizing profile items to broader categories, the approach changes the state of profile items by adding noise and obfuscation. The original granular profile information is preserved but transformed into a noisy version that maintains precision while preventing direct identification
4Object-affected harmful factors
If users disable web cookies or browse in private mode to limit tracking, then user privacy is improved, but personalization capability is lost
Solution Approach 1:
Instead of disabling cookies to protect privacy, the approach inverts the problem by creating a special type of cookie that inherently protects privacy while enabling personalization. The privacy-preserving cookie allows users to keep cookies enabled without sacrificing privacy
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The privacy-preserving cookie generator implementations described herein create a privacy-preserving data structure (also called a privacy-preserving cookie herein) that is used to provide personalization for online services without user tracking. In some implementations the privacy-preserving cookie generator encodes a user profile (for example, based on a user's online activity) into a data structure that has naturally occurring noise and that efficiently supports noise addition. In one implementation a Bloom filter is used to create the encoded profile. Additional noise is injected into the encoded profile to create an obfuscated user profile in the form of a privacy-preserving data structure. The privacy-preserving data structure or cookie can be attached to an online service request and sent over a network to an online service provider which can use it fulfill the services request, providing a somewhat personalized result while the user's privacy is maintained.