Native Content Asset Rendering With Predicted Slot Size
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately predict the size of publisher-rendered native content items in applications, leading to inefficiencies and degradation in image quality due to insufficient utilization of user interface display space.
Innovation Solution
A model is developed to predict the size of publisher-rendered native content item slots and generate a replacement main asset, using a size hint pipeline and content element generation pipeline, which includes a size hint model and a content element generation model, to improve the display of publisher-rendered native content items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing systems are used to display publisher-rendered native content items, then the content can be displayed, but the image quality degrades and display space is underutilized due to inaccurate size prediction
Solution Approach 1:
The system performs preliminary size prediction using a size hint model before generating the main asset. This allows the asset generation process to proceed with accurate size information, preventing image quality degradation that would occur if assets were generated with incorrect size assumptions. The model predicts size based on historical data and content characteristics before the actual rendering occurs.
Solution Approach 2:
The system implements a feedback mechanism where actual rendered content sizes are fed back to update and refine the size hint model. This continuous learning process improves prediction accuracy over time, allowing the system to adapt to different content types and publishing patterns, thereby maintaining high image quality across diverse content scenarios.
2Area of stationary object
If existing systems display publisher-rendered native content items, then content presentation is possible, but negative space increases and display space utilization decreases
Solution Approach 1:
The system calculates the expected size of publisher-rendered content items before generating assets using the size hint model. This preliminary size information enables precise asset generation that fills the available display space without creating unnecessary negative space, maximizing the utilization of stationary display areas.
Solution Approach 2:
The system dynamically adjusts asset generation parameters based on the predicted size characteristics of publisher-rendered content. By changing parameters such as asset dimensions, resolution, and composition based on the expected rendering size, the system eliminates negative space and ensures optimal display space utilization while preserving all content information.
Data Source
AI summary
Example embodiments of the present disclosure provide for an example method including obtaining a request for third party content elements to be used in a publisher-rendered native content item slot. The example method includes obtaining input signals comprising data indicative of publisher-rendered native content items. The example method includes generating, by a model associated with the content provider computing system, based on the input signals, an expected native content item slot size. The example method includes obtaining, by the content provider computing system, content item elements from a content provider. The example method includes generating based on the expected native content item slot size, a replacement main asset comprising a first content element of the one or more content elements. The example method includes transmitting data including the replacement main asset and additional content elements to a publisher to be used in a publisher-rendered native content item slot.