User Behavior Indicator for Video Segment Tagging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face challenges in identifying relevant or interesting content within long videos, as they must manually navigate through less interesting segments to find valuable material, leading to wasted time and reduced content sharing in social networks.
Innovation Solution
A system that partitions videos into segments, monitors user behavior, and logs interactions to determine normal behavior patterns, tagging segments that deviate from these patterns for display as indicators of user interest, allowing users to quickly find preferred content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If users manually navigate through long videos to find interesting content, then they can discover relevant material, but they waste significant time on irrelevant segments
Solution Approach 1:
The video is divided into multiple segments, and user behavior is analyzed at the segment level rather than for the entire video. This allows the system to identify specific interesting segments without requiring users to watch the whole video, thereby reducing time loss while maintaining ease of navigation.
Solution Approach 2:
The system pre-analyzes user behavior patterns by monitoring how users interact with video segments before the current user views the video. These pre-computed behavior indicators are then used to guide the current user, eliminating the need for manual navigation through irrelevant content.
2Measurement precision
If the system monitors and analyzes user behavior in detail, then it can accurately identify interesting segments, but the system complexity increases
Solution Approach 1:
The system extracts only the essential user behavior data needed for analysis (such as pause duration, replay count, skip behavior) and ignores irrelevant details. This extraction approach maintains measurement precision while reducing system complexity by focusing only on critical behavior indicators.
Solution Approach 2:
The system uses the users' own behavior data to generate the behavior indicators, eliminating the need for external annotation or manual labeling. Users inadvertently provide the data through their natural viewing patterns, reducing system complexity while maintaining high accuracy.
3Ease of operation
If the system tags and displays behavior indicators for all segments, then users can easily find interesting content, but the information overload increases
Solution Approach 1:
Instead of providing uniform information for all video segments, the system applies different levels of information display based on local segment characteristics. Segments with high user interest receive prominent indicators, while less interesting segments receive minimal or no indicators, reducing information overload while maintaining ease of content discovery.
Solution Approach 2:
The system displays behavior indicators selectively for only the most interesting segments rather than for all segments. This partial action approach prevents information overload by showing users only the most relevant parts of the video, while still providing enough information to easily discover interesting content.
Data Source
AI summary
Techniques for determining user behavior related to media are described. A media file containing media is received. The media is partitioned into segments. The user-interaction analyzer monitors user behavior with respect to viewing the media and the segments. The user's behavior with respect to viewing the segments is logged. Normal user behavior with respect to the media is determined and stored. Logged user behavior with respect to interaction with a segment of the media is compared with the normal user behavior with respect to the media. Logged user behavior of a particular media segment that deviates from normal relative to the determined normal user behavior is determined. A particular media segment that is not normal is tagged. A user device is configured to display indicia indicating user behavior related to media to help inform which parts users tend to like the most.