Image Viewing History Analysis With Shapley Contribution Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image distribution providers face challenges in identifying which index values contribute to deteriorating experience quality, making it difficult to design or redesign systems to improve user satisfaction.
Innovation Solution
An information processing apparatus that calculates a contribution degree of each index value in viewing history data using cooperative game theory, specifically through Shapley values, to determine the impact of each index on experience quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a complicated experience quality estimation model with multiple index values is used, then the accuracy of experience quality estimation is improved, but the ability to identify which index value causes deterioration in experience quality deteriorates
Solution Approach 1:
The patent segments the complex experience quality estimation model into individual index value contributions by calculating the contribution degree of each index value separately. This allows the system to maintain the accuracy of the comprehensive model while identifying which specific index values (such as bit rate, resolution, frame rate) contribute most to experience quality deterioration, thereby resolving the contradiction between model accuracy and identifiability of contributing factors.
2Adaptability or versatility
If multiple index values are input to the estimation model, then the comprehensiveness of experience quality assessment is improved, but the complexity of the analysis process increases
Solution Approach 1:
The patent introduces a feedback mechanism where the contribution degree calculation unit provides quantitative feedback about each index value's impact on experience quality. This feedback enables the system to comprehensively assess multiple index values (bit rate, resolution, frame rate, etc.) while simplifying the analysis process by automatically ranking their contributions, thus resolving the contradiction between comprehensiveness and analytical complexity.
Data Source
AI summary
An information processing apparatus including: an experience quality estimation unit that estimates an experience quality value when changing any index value of a plurality of index values included in viewing history data of an image; a contribution degree calculation unit that calculates a contribution degree of each index value included in the plurality of index values to a contribution degree on the basis of the estimated experience quality value; and a contribution degree output unit that outputs information indicating the contribution degree.