Predicting User Characteristics for Online Content Targeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Online systems face challenges in selecting relevant content for third-party system users due to missing user characteristics, such as age, which are not maintained by the third-party systems, leading to inaccurate content presentation and reduced user interaction.
Innovation Solution
The online system predicts missing characteristics, like age, by generating models based on obtained characteristics from a sampled user set and adjusts accuracy using verified data from trusted third-party systems, ensuring more accurate content targeting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the online system uses only the characteristics maintained by the third party system to select content, then the content selection process is simple, but the content selection accuracy deteriorates due to missing user characteristics
Solution Approach 1:
The online system acts as an intermediary that receives characteristics from the third party system, predicts additional characteristics using machine learning models, and combines both sources of information to make content selection decisions. This mediator approach allows the system to leverage the simplicity of third party data while adding the precision of predicted characteristics.
Solution Approach 2:
The solution segments the characteristic data into two distinct sources: characteristics directly maintained by the third party system and characteristics predicted by the online system using machine learning models. This segmentation allows each source to be processed and validated independently, improving overall accuracy while maintaining process clarity.
2Measurement precision
If the online system predicts characteristics for all third party system users, then the content targeting accuracy is improved, but the computational resources and time required increase
Solution Approach 1:
The system implements partial action by predicting characteristics only for users where such prediction is necessary and beneficial for content selection, rather than universally for all users. The system can choose to use predicted characteristics selectively based on content requirements, third party system capabilities, and user-specific needs, thereby reducing unnecessary computational overhead.
3Measurement precision
If the online system obtains verified characteristics from trusted third party systems, then the accuracy of predicted characteristics is improved, but the system complexity and data integration requirements increase
Solution Approach 1:
The online system serves as an intermediary that selectively integrates verified characteristics from trusted third party systems. It maintains a curated list of trusted sources and selectively incorporates their data into the prediction and selection process, managing integration complexity through controlled partnerships rather than universal integration.
Data Source
AI summary
An online system maintains characteristics for its users and may access characteristics of users maintained by a third party system. The online system may select content for a user of the third party system based on characteristics maintained by the third party system. If the third party system does not maintain a characteristic for its users, the generates a model predicting the characteristic for third party system users based on a set of online system users identified based on characteristics of third party system users. The online system clusters third party system users based on the predicted characteristic for other third party system users connected to the third party system user. Using verified characteristics for third party system users from a trusted third party system, the online system determines an accuracy of the predicted characteristic for third party system users in a cluster.