Encrypted Ad Templates for Privacy-Preserving Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional advertising models struggle to provide highly personalized advertisements while maintaining user privacy and confidentiality, as they rely on shared user data that may raise security concerns and limit ad targeting effectiveness.
Innovation Solution
A method for generating dynamically customized user-specific advertisements using an ad template with functions that incorporate personalized user information, where the data is stored as key-value pairs to obscure meaning from the serving entity, allowing for detailed targeting without compromising privacy, and enabling multiple entities to contribute personalized content without revealing their proprietary information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If personalized user information is shared between advertisers and social networking systems for ad targeting, then ad personalization effectiveness is improved, but user privacy and data security are compromised
Solution Approach 1:
The patent introduces encrypted ad templates as an intermediary mechanism that enables personalized advertising without direct sharing of sensitive user data. The encryption system acts as a mediator that allows the social networking system to serve personalized ads while keeping user information confidential, thus resolving the contradiction between ad personalization and privacy protection.
Solution Approach 2:
The patent transforms user information from its original readable form into encrypted parameter representations. By changing the state of data from plaintext to encrypted format, the system maintains the functional ability to personalize ads while eliminating the privacy risk associated with exposing raw user information.
2Productivity
If advertisers disclose proprietary campaign information to social networking systems for ad serving, then ad delivery capability is improved, but advertiser confidentiality is compromised
Solution Approach 1:
Encrypted ad templates serve as an intermediary that enables the social networking system to deliver ads effectively without requiring advertisers to disclose proprietary campaign information in readable form. The encryption mechanism allows the system to process and deliver ads while maintaining advertiser confidentiality.
Solution Approach 2:
The patent creates encrypted copies of ad templates that contain all necessary information for ad delivery but are protected from direct inspection. These encrypted copies allow the social networking system to perform its ad serving function while the original proprietary information remains confidential to the advertiser.
3Measurement precision
If detailed user data is collected and stored for personalized advertising, then ad targeting precision is improved, but system complexity and security risks increase
Solution Approach 1:
The patent changes the state of user data from plaintext to encrypted format, which maintains the precision needed for targeting while reducing the security risks and complexity associated with storing and managing sensitive readable data. The encrypted parameters retain all necessary information for precise targeting without exposing user details.
Data Source
AI summary
An entity such as a social networking system generates a dynamically customized user-specific advertisement using an advertisement template and user information. A request is received for an advertisement to display to a user, and the social networking system retrieves an advertisement template that includes one or more functions. The social networking system also retrieves user information to evaluate a value of a function contained in the advertisement. The user information is provided by both a first and second party (e.g., the advertiser and the social networking system), where the meaning of the information provided by each party is obscured from the other party. The template and the value of each function are used to generate a dynamically customized user-specific advertisement for display to the user, where both parties have provided user information that is obscured in some manner from the other party, allowing for a more personalized advertisement.