Neural Attribution Model for Content Impact Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for attributing user actions to content presentation operations rely on heuristics and linear models that fail to account for interactions and timing, leading to suboptimal content presentation plans and resource inefficiencies.
Innovation Solution
A deep neural network-based attribution model using convolutional layers and attention mechanisms to analyze user interaction sequences, providing data-driven estimates of the influence of content presentation operations on user actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If heuristics and linear models are used for attributing user actions to content presentation operations, then the implementation is simple, but the attribution accuracy is low and fails to account for interactions and timing
Solution Approach 1:
The patent replaces conventional heuristics and linear models with a deep neural network-based attribution model. This substitution transitions from simple mechanical calculation methods to an intelligent system capable of capturing complex interactions and timing effects, thereby significantly improving attribution accuracy while accepting increased model complexity.
Solution Approach 2:
The attribution model combines multiple components including convolutional layers for feature extraction, attention mechanisms for weighting interactions, and integration with content presentation operation data. This composite structure enables the system to simultaneously process multiple factors (interactions, timing, user behavior) that single-method approaches cannot capture, resolving the contradiction between accuracy and complexity.
2Measurement precision
If deep neural network-based attribution model is used, then the attribution accuracy improves by accounting for interactions and timing, but the resource consumption increases
Solution Approach 1:
The deep neural network is divided into specialized modules: convolutional layers for extracting features from user interaction sequences, attention mechanisms for identifying important interactions, and separate processing streams for different types of data. This segmentation allows efficient parallel processing and reduces redundant computations, lowering resource consumption while maintaining high attribution accuracy.
Solution Approach 2:
The system performs preliminary processing of user interaction data through convolutional layers to extract relevant features before passing them to the attention mechanism. This preliminary action filters and prepares data in advance, reducing the computational burden on subsequent stages and optimizing overall resource utilization while preserving attribution accuracy.
3Productivity
If conventional linear models are used for content presentation planning, then the resource consumption is low, but the content presentation plans are suboptimal
Solution Approach 1:
The attribution model provides feedback by accurately measuring the impact of each content presentation operation on user actions. This feedback loop enables the system to identify which operations are effective and which are not, allowing for continuous optimization of content presentation plans. The feedback mechanism transforms resource spending from a blind process to a data-driven decision-making process, improving productivity while reducing waste.
Solution Approach 2:
The system transitions from static content presentation plans based on linear models to dynamic plans that adapt based on real-time attribution data. The deep neural network continuously processes new interaction data, updates attribution estimates, and recommends optimized content presentation strategies. This dynamic approach maximizes productivity by constantly aligning resource allocation with actual user response patterns.
Data Source
AI summary
Interaction data sets of consumers of an offering set with an application associated with an offering set are obtained. At least some of the interactions are associated with respective content presentation operations. A neural network model is trained, using the interaction data sets, to generate predictions of actions of a consumer at whom a set of content presentation operations is directed, and to provide action-attribution scores for the content presentation operations. The model includes convolutional layers and an attention module. Weights learned in the attention module are used to estimate the action-attribution scores. A trained version of the model is stored.