AI Viewership Prediction for Content Delivery Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media content delivery systems face inefficiencies in utilizing limited resources to maximize viewer interest while minimizing unwatched content, and there is a need for accurate consumer behavior analysis with minimal user inconvenience.
Innovation Solution
A system that monitors consumer behavior through viewership data, determines missing data points, and uses artificial intelligence to predict substitute data points, enabling detailed analysis across various scopes (in-episode, cross-episode, cross-season, in-network, and cross-network) to optimize content delivery and advertising.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user surveys are used to determine consumer behaviors and intents, then accuracy of consumer behavior data is improved, but user convenience deteriorates due to inconvenience to users
Solution Approach 1:
The system automatically collects viewership data through set-top boxes and monitoring devices without requiring user participation in surveys. The consumer behavior data is gathered passively as users naturally watch content, eliminating the need for users to complete survey forms or provide explicit feedback, thus maintaining both data accuracy and user convenience.
Solution Approach 2:
The patent introduces an intermediary monitoring system between the user and the content delivery system. This intermediary automatically captures viewership data through set-top boxes and behavioral tracking, serving as a mediator that collects accurate consumer behavior information without directly involving or inconveniencing the users.
2Adaptability or versatility
If resources are allocated to transmit all available media content, then content availability is improved, but resource efficiency deteriorates due to limited transmission bandwidth and routing devices
Solution Approach 1:
The system performs preliminary analysis of consumer behavior data and viewership patterns before allocating transmission resources. By predicting which content is most likely to be watched based on historical data and user profiles, the system pre-configures resource allocation to prioritize high-demand content, ensuring both content availability and resource efficiency.
Solution Approach 2:
The patent dynamically changes transmission resource allocation parameters based on real-time viewership data and predicted consumer behavior. The system adjusts bandwidth allocation, routing priorities, and content delivery timing according to measured usage patterns, optimizing resource efficiency while maintaining content availability for watched programs.
3Measurement precision
If detailed viewership data is collected at highly granular levels, then analysis precision is improved, but data processing complexity deteriorates
Solution Approach 1:
The patent segments viewership data collection into distinct granular levels (e.g., program-level, episode-level, scene-level) and processes each segment separately. The system collects detailed data only where necessary for specific analysis objectives, reducing overall processing complexity while maintaining high analysis precision for targeted measurements.
Solution Approach 2:
The system collects detailed granular viewership data selectively for specific programs or time periods when analysis precision is most needed, rather than uniformly collecting maximum detail for all content. This partial action approach maintains high analysis precision for priority measurements while reducing overall data processing complexity.
Data Source
AI summary
A method is described for obtaining and using viewership data to determine relationships between previous viewership of an episodic program and viewership of future episodes of the program. Aggregating the data may comprise determining missing data points and predicting substitute data points based on past viewership data for the individual viewer and a plurality of other viewers. Data may be presented to the user in a number of ways to aid in analysis and planning of likely viewership for a season of a television program. Data may be used to optimize advertising revenues, plan television lineups to maximize a number of likely viewers, or allocate content and resources between broadcast and video-on-demand.