Re-Watch Content Suggestions Using Time-to-Forget Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional content streaming systems fail to suggest re-watching content when users are likely to have forgotten it, instead recommending recently viewed content without considering the user's likelihood of forgetting.
Innovation Solution
A computer-implemented method utilizing a TTF threshold value and a consumption score (CS) to suggest re-watching content after the user is likely to have forgotten it, incorporating an AI model to adjust these values based on viewing habits and preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional content streaming platforms provide content suggestions, then users can discover new content, but the suggestions do not account for users having forgotten previously viewed content, reducing suggestion effectiveness
Solution Approach 1:
The patent introduces a Time to Forget (TTF) threshold as an intermediary metric that mediates between the system's content history data and the user's actual memory state. This TTF threshold serves as a computational model to estimate when users have forgotten previously viewed content, enabling the recommendation system to accurately identify forgotten content without directly accessing user memory.
Solution Approach 2:
The patent dynamically adjusts the TTF threshold parameter based on multiple factors including consumption score, series time, and days since last viewed. By changing this parameter adaptively, the system optimizes its ability to identify forgotten content across different user behaviors and content types, thereby improving suggestion accuracy while accounting for memory loss.
2Measurement precision
If the platform tracks detailed content history to improve suggestions, then suggestion accuracy improves, but system complexity increases
Solution Approach 1:
The patent segments the complex problem of identifying forgotten content into distinct computational components: consumption score calculation, series time analysis, days since last viewed tracking, and TTF threshold determination. Each component processes specific aspects of content history independently, then integrates results to produce accurate re-watch suggestions without requiring a monolithic complex system.
Solution Approach 2:
The system performs preliminary calculations of consumption scores and TTF thresholds in advance, storing these pre-computed values for efficient retrieval during recommendation generation. This preliminary action reduces real-time computational complexity while maintaining high accuracy in identifying forgotten content for re-watch suggestions.
Data Source
AI summary
Embodiments determine a user who is watching current content, receive content history of the user, calculate a consumption score (CS) based on series time, watch time, and consumption time of the content history, calculate a time to forget (TTF) threshold value based on the CS, the series time, and days since last viewed of the content history, compare the CS and the TTF threshold value to the current content watched by the user and current viewing habits of the user, and provide at least one suggestion to re-watch content of the content history based on the comparing.