Machine-Learning Cell Twinning for Unbiased Content Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face errors in determining the performance of digital content due to differences in user profiles and behaviors between exposed and control cells, leading to bias and resource-intensive compensation algorithms.
Innovation Solution
Utilize machine learning to generate a parameter ranking and construct a digital twin of a cell of computing devices that accessed digital content, emphasizing similarity in demographics and behavior to reduce bias and resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional control cells are used to determine digital content performance, then performance measurement can be conducted, but errors arise due to differences in user profiles and behaviors between exposed and control cells
Solution Approach 1:
The patent creates a digital twin of the control cell that copies the demographic and behavioral characteristics of the exposed cell. This twin is generated using machine learning models that replicate user profiles, device attributes, and interaction patterns, enabling accurate performance measurement without the biases inherent in traditional control cells
Solution Approach 2:
The patent transforms the control cell parameters by applying machine learning-derived adjustments based on the exposed cell's actual characteristics. Instead of using a static or randomly generated control cell, the system dynamically adjusts parameters such as user demographics, device types, and behavioral patterns to match the exposed cell, thereby eliminating measurement errors
2Measurement precision
If compensation algorithms are used to correct biases in control cells, then measurement accuracy can be improved, but processing resources and power consumption increase
Solution Approach 1:
The patent performs preliminary action by pre-generating a digital twin of the control cell using machine learning models before the performance measurement process begins. This twin is created in advance with all necessary demographic and behavioral characteristics already matched to the exposed cell, eliminating the need for resource-intensive compensation algorithms during the actual measurement process
Solution Approach 2:
The patent replaces the mechanical system of iterative compensation algorithms with a machine learning-based digital twin generation system. Instead of continuously adjusting and compensating for biases during measurement, the system uses trained ML models to directly generate a matched control cell representation, significantly reducing computational overhead and power consumption
Data Source
AI summary
Twinning is disclosed based on parameter ranking. A system can include a data processing system. The data processing system can include one or more processors, couple with memory. The data processing system can receive identifiers of computing devices of a cell that accessed digital content. The data processing system can receive parameters linked with the identifiers. The data processing system can determine, based on the parameters and a model trained with machine learning, a ranking of the parameters. The data processing system can generate, based on the ranking of the parameters, a twin of the cell comprising identifiers of computing devices that did not access the digital content.