Video Recommendation via Affect Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating video responses are imprecise, subjective, and unreliable, particularly when assessing brief periods of video content, leading to ineffective video recommendations.
Innovation Solution
A computer-implemented method that captures mental state data, including physiological and facial data, while playing a video, and recommends subsequent media presentations based on aggregated data from individuals and groups, using correlation analysis to infer emotional and cognitive states and provide personalized recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional star rating methods are used for video evaluation, then the evaluation process is simple to implement, but the precision and reliability of recommendations are poor
Solution Approach 1:
The patent replaces manual star rating mechanisms with automated affect recognition technology that captures physiological data (heart rate, skin conductance, respiration) and facial expressions to objectively measure viewer emotional responses, eliminating subjective human input while significantly improving evaluation precision
Solution Approach 2:
The system introduces affect recognition software and physiological sensors as intermediary components between the viewer and the recommendation engine, capturing emotional states during video playback and translating them into quantifiable data that drives personalized recommendations
2Productivity
If self-rating methods are used, then users can provide feedback, but the process is tedious and time-consuming
Solution Approach 1:
The system enables self-service evaluation by automatically capturing and analyzing the viewer's own physiological responses and facial expressions during video playback, eliminating the need for manual rating input while still gathering comprehensive feedback data
Solution Approach 2:
The affect recognition system operates continuously throughout the entire video playback process, capturing emotional responses in real-time without interrupting the viewing experience or requiring the user to pause for evaluation input
3Measurement precision
If brief video portions are evaluated, then specific segments can be assessed, but the evaluation becomes more difficult and complex
Solution Approach 1:
The system automatically segments the video into discrete portions and evaluates affect responses for each segment independently, allowing precise identification of which specific video portions elicited particular emotional reactions without requiring manual segmentation effort
4Reliability
If traditional recommendation systems are used, then recommendations can be provided, but they are imprecise and often unreliable
Solution Approach 1:
The system implements continuous feedback loops where affect data collected from viewers is fed back into the recommendation algorithm, allowing the system to learn and adapt to individual viewer preferences and emotional patterns, significantly improving recommendation reliability over time
Solution Approach 2:
The system performs preliminary affect analysis during the initial video playback to establish baseline viewer preferences and emotional responses before generating personalized recommendations, ensuring that recommendations are based on actual measured data rather than generic algorithms
Data Source
AI summary
Analysis of mental states is provided to enable data analysis pertaining to video recommendation based on affect. Video response may be evaluated based on viewing and sampling various videos. Data is captured for viewers of a video where the data includes facial information and/or physiological data. Facial and physiological information may be gathered for a group of viewers. In some embodiments, demographics information is collected and used as a criterion for visualization of affect responses to videos. In some embodiments, data captured from an individual viewer or group of viewers is used to rank videos.