Privacy-Preserving Content Selection via Local Attribute Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in selecting relevant content for users without compromising user privacy or data security, especially with the deprecation of third-party cookies.
Innovation Solution
The solution involves predicting user attribute buckets relevant to a user based on contextual data and updating user attribute information stored on the client device, allowing content platforms to select relevant content without tracking users across multiple resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user tracking across multiple resources is implemented to select relevant content, then content relevance is improved, but user privacy is compromised
Solution Approach 1:
The patent segments user attribute information into platform-specific attribute buckets stored locally on each client device. Each content platform maintains its own separate user attributes without cross-platform tracking, dividing the user profile management into isolated segments that cannot be combined to identify users across different resources.
Solution Approach 2:
The patent introduces an intermediary mechanism where machine learning models predict user attributes based on contextual data from individual platform interactions. These predictions serve as intermediaries between user behavior and content selection, enabling relevant content delivery without directly tracking or storing sensitive user information across platforms.
2Loss of information
If third-party cookies are used for user tracking, then cross-site data collection is improved, but data security is compromised
Solution Approach 1:
The patent extracts the essential function of user tracking by taking out only the necessary user attribute predictions needed for content selection. Instead of using third-party cookies to collect comprehensive user data, the system extracts and stores only platform-specific attribute predictions locally on the client device, eliminating the security vulnerabilities of cross-site tracking while maintaining content relevance.
3Object-affected harmful factors
If user attribute information is stored on the client device, then user privacy is improved, but network bandwidth consumption increases
Solution Approach 1:
The patent applies preliminary action by pre-storing user attribute predictions and content selection criteria on the client device before actual content requests. The machine learning models generate user attribute predictions in advance based on contextual data, and these predictions are cached locally, enabling rapid content selection without repeated network round trips for each content request.
4Measurement precision
If machine learning models predict user attributes based on contextual data, then content selection accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies partial action by using machine learning models to predict only the specific user attributes necessary for content selection rather than analyzing all possible user characteristics. The system focuses computational resources on predicting relevant attributes like topic interest and content preference based on contextual data, avoiding unnecessary computational overhead from comprehensive user profiling.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, systems, and apparatus, including medium-encoded computer program products for selecting and displaying content in privacy preserving manners are described. A digital component request that includes contextual data related to an environment in which the digital component will be displayed can be received from a client device by a first content platform. Based on the contextual data, the user can be assigned to user attribute buckets, which can be associated with at least one type of user attribute. Based on the contextual data and each user attribute bucket to which the user is assigned, candidate digital components can be selected for distribution to the client device. The application can be provided response data that initiates an update to aggregated user attribute data for the user and enables the application to select a digital component to display to the user based on the aggregated data.