Social Network Content Delivery via Relationship Proximity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current statistical models used for predicting user behavior and content recommendations in online environments are imperfect, leading to less relevant product suggestions for sellers, advertisers, and web site operators.
Innovation Solution
A method for selecting relevant online content based on the prior online activities of members within a social network, considering the closeness of relationships and click-through rates within predefined groups, with the option to prioritize ads with high revenue-generating potential and deliver ads based on predicted click probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If statistical models based on collaborative filtering techniques are used to predict user behavior, then product recommendations can be generated, but the accuracy and relevance of recommendations are insufficient
Solution Approach 1:
The patent introduces social network relationships as an intermediary layer between user behavior data and recommendation generation. By incorporating friendship connections and social proximity metrics, the system mediates the prediction process to achieve more accurate and reliable recommendations that reflect both individual behavior and social influence patterns
Solution Approach 2:
The patent transforms the prediction parameters by adding social network parameters (friendship connections, degree of separation, social proximity scores) to the traditional collaborative filtering parameters. This parameter expansion allows the system to capture more dimensions of user behavior and improve prediction accuracy while maintaining computational feasibility
2Productivity
If online ads are selected based on high per-click revenue, then revenue-generating potential is maximized, but the relevance to user interests may be reduced
Solution Approach 1:
The patent changes the ad selection parameters by incorporating social network-based relevance metrics alongside revenue metrics. The system calculates social proximity scores and uses them to weight ad relevance, creating a balanced selection criterion that considers both revenue potential and user interest alignment rather than relying solely on per-click revenue
Solution Approach 2:
The patent implements a feedback mechanism where ad performance data (click-through rates, user interactions) is continuously collected and used to refine both revenue estimates and relevance predictions. This feedback loop allows the system to learn from actual user behavior and adjust ad selections to optimize both revenue and relevance over time
3Quantity of substance
If content delivery is based on broad demographic groups, then coverage is maximized, but the precision of targeting is reduced
Solution Approach 1:
The patent segments the audience using social network relationships, dividing users into groups based on their proximity to the target user (direct friends, friends of friends, etc.). This segmentation enables the system to maintain broad coverage across multiple social layers while achieving precise targeting within each segment by leveraging the homophily principle that users tend to share interests with their social connections
Data Source
AI summary
Relevant content is prepared and selected for delivery to a member of a network based, in part, on prior online activities of the other members of the network, and the closeness of the member's relationship with the other members of the network. The relevant content may be an online ad, and is selected from a number of candidate online ads based on click-through rates of groups that are predefined with respect to the member and with respect to certain attributes. An online ad's revenue-generating potential may be considered in the selection process.