Generative Ad Creation Using Conditional Media Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current online advertising methods are resource-intensive and costly due to the manual creation and selection of online ads, which limits efficiency and customization in presenting ads to users based on their profiles and interactions.
Innovation Solution
The use of generative processes, such as conditional generative processes trained with user-interaction and prototype media content, to create non-pre-existing media objects that appeal to users and align with advertiser messaging, allowing for real-time or near real-time ad creation and personalization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If manual creation and selection of online ads is used, then ad quality and relevance to user profiles can be maintained, but resource consumption and costs increase significantly
Solution Approach 1:
The system enables self-service by training generative processes on user-interaction content and prototype media content, allowing the system to automatically create media objects that appeal to specific users without manual intervention. The generative process serves itself by learning from historical data and autonomously generating personalized ad content.
Solution Approach 2:
The patent replaces the mechanical system of manual ad creation with a computational generative process. Instead of human designers manually creating ads, a trained generative model automatically generates media objects by processing user profiles and interaction histories, substituting human labor with an automated AI system.
2Productivity
If manual ad design processes are used, then customization to user profiles is possible, but efficiency and speed of ad creation decrease
Solution Approach 1:
The system performs preliminary action by training the generative process in advance on extensive user-interaction content and prototype media content. This pre-training enables the model to quickly generate customized media objects during actual ad creation without requiring manual design for each user, thus improving efficiency while maintaining customization capability.
Solution Approach 2:
The generative process dynamically changes parameters based on user profiles and interaction histories. By adjusting the generation parameters according to specific user characteristics, the system efficiently produces customized media objects tailored to each user's preferences and behaviors, maintaining adaptability while automating the process.
3Adaptability or versatility
If generative processes are used to create media objects, then customization and efficiency improve, but system complexity increases
Solution Approach 1:
The generative process serves multiple functions: it learns from user-interaction content, processes prototype media content, generates customized media objects, and adapts to different user profiles. This multi-functionality consolidates what would otherwise require multiple separate systems into a single versatile generative model, managing complexity through consolidation.
4Loss of time
If real-time or near real-time ad creation is implemented, then user engagement increases, but computational resources required increase
Solution Approach 1:
The system performs preliminary training of the generative process on extensive datasets before real-time operation. This pre-computation allows the model to make quick inferences during actual ad creation, reducing real-time computational requirements while enabling fast, customized ad generation that improves user engagement.
Data Source
AI summary
Embodiments of methods and/or systems for constructing one or more online ads using one or more media objects created using one or more generative processes are disclosed.