Video Ranking via Lead Video Session Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video distribution systems lack efficient methods to rank and present videos based on user engagement, often recommending popular but not necessarily relevant videos, which can lead to increased search time and data transfer for users.
Innovation Solution
The system identifies lead videos that initiate presentation sessions and applies scaling factors to determine presentation scores, ranking videos based on their contribution to session time, with higher scores for videos that bring users to the platform or are shared through social channels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If videos are recommended based on popularity, then video distribution is simplified, but user relevance and search time are worsened
Solution Approach 1:
The system performs preliminary analysis of presentation session data to identify lead videos before recommending them to users. By pre-calculating presentation scores based on historical session patterns, the system can quickly retrieve and present relevant videos without requiring users to perform extensive searches, thus reducing search time while maintaining recommendation simplicity
Solution Approach 2:
The system uses feedback from presentation session data to continuously refine video recommendations. By analyzing which videos initiate sessions and how long users engage with them, the system adjusts its ranking algorithms to prioritize videos with higher presentation scores, thereby reducing the time users spend searching for relevant content
2Device complexity
If all videos are presented equally, then system complexity is reduced, but user engagement quality is worsened
Solution Approach 1:
The system applies different quality treatment to different videos based on their presentation scores. Instead of uniform treatment, videos are segmented into categories (lead videos with high scores versus other videos), and each category receives appropriate weighting in the recommendation algorithm. This local differentiation improves relevance without requiring complete system redesign
Solution Approach 2:
The system changes the parameter of video weighting based on presentation scores. Videos with higher presentation scores receive increased weight in the recommendation calculation, while others receive standard weight. This parameter adjustment allows the system to maintain relatively simple architecture while significantly improving video relevance to user preferences
3Productivity
If lead videos are prioritized with scaling factors, then video distribution efficiency is improved, but calculation complexity is worsened
Solution Approach 1:
The system applies scaling factors selectively only to lead videos rather than all videos. By identifying lead videos through session start data and applying enhanced weighting only to this subset, the system achieves improved distribution efficiency for relevant content without the computational burden of complex calculations across the entire video library
Solution Approach 2:
The video library is segmented into lead videos and other videos based on presentation session analysis. This segmentation allows the system to apply different calculation rules to different segments, improving efficiency by focusing computational resources only where needed (on lead video identification and scoring) rather than uniformly processing all videos
Data Source
AI summary
Systems, methods, and apparatus include computer programs encoded on a computer-readable storage medium, including a system for ranking videos. Videos are identified that have been presented at client devices. For each video, session start data is identified that specifies a lead video that initiated presentation to a user during a presentation session. For each lead video, presentation times over multiple user sessions are determined, a scaled presentation time is obtained, user sessions for which the lead video initiated presentation of videos are identified, and an aggregate video presentation time attributable to the lead video is determined. For each given video, a presentation score is determined based on a scaled presentation time of the lead video relative to a sum of the aggregate video presentation times for the lead videos. The videos are ranked based on the presentation scores. A user interface is updated to present the ranked videos.