Digital Component Selection Using Predicted Display Position
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing content distribution systems struggle to efficiently select and display digital components, particularly non-observable parameters such as the vertical display position of digital components, which can significantly impact the effectiveness of display, but are not known to the system until the actual display position, which can affect the display of digital components in electronic documents.
Innovation Solution
A content distribution system uses a parameter prediction model to predict the value of non-observable parameters, such as the vertical display position, by processing observable parameters available in the request, and then uses these predictions to select digital components that are better adapted for display, reducing latency and improving relevance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system waits to obtain actual non-observable parameters (such as vertical display position) before selecting digital components, then the selection accuracy is improved, but the latency increases
Solution Approach 1:
The system performs preliminary prediction of non-observable parameters (such as vertical display position) using a parameter prediction model before the actual display occurs. This allows the digital component selection to be made in advance based on predicted values, resolving the contradiction by obtaining accurate parameter estimates beforehand without waiting for actual observation, thus maintaining selection accuracy while reducing latency
2Loss of time
If the system uses predicted non-observable parameters for digital component selection, then the latency is reduced, but the measurement precision deteriorates
Solution Approach 1:
The system employs feedback mechanisms where the predicted non-observable parameters are continuously refined based on actual observed values when they become available. The parameter prediction model is trained and updated using historical data and actual outcomes, improving the accuracy of predictions over time. This feedback loop ensures that while predictions are made early to reduce latency, the precision is maintained and improved through iterative learning from actual data
Data Source
Figure 1
Figure 2A~2C
Figure 3
AI summary
Systems, methods, devices, and techniques for improving the efficiency of selecting digital components to present in electronic documents and reducing latency in rendering digital components in electronic documents. In some implementations, a content distribution system uses predicted metrics for a set of candidate components to determine items to present in an electronic document responsive to a request. A metric prediction model can generate predicted metrics with the aid of a parameter prediction sub-model that predicts a value of a non-observable parameter associated with a request for a digital component.