Video Parameter Selection via KPI Significance Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video distribution service technologies fail to optimize moving-image parameters in real-time to enhance user satisfaction and profit, as they are based on limited QoE estimation models that do not account for factors like price and content, and are not adaptable to varying conditions.
Innovation Solution
A moving-image parameter selection device that associates video distribution parameter combinations with Key Performance Indicators (KPIs), calculates averages, and performs significance tests to select optimal parameter combinations for improved distribution control, considering past data and user feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a QoE estimation model based on subjective evaluation experiments is used, then moving-image parameters can be controlled to optimize QoE, but the model is limited to only examining QoE caused by moving-image quality or playback quality and does not account for other user satisfaction factors such as price and content
Solution Approach 1:
The patent changes the fundamental parameter being measured from QoE (Quality of Experience) to KPI (Key Performance Indicator) representing actual user satisfaction and profit metrics. This allows the system to evaluate moving-image parameters based on comprehensive business outcomes including price, content, user satisfaction, and profit, rather than being limited to technical quality metrics only.
Solution Approach 2:
The patent implements a feedback mechanism where actual KPI data from video distributions is collected, stored, and used to evaluate and select optimal moving-image parameter combinations. This closed-loop feedback system enables continuous optimization based on real performance data rather than theoretical QoE models.
2Reliability
If moving-image parameters are optimized based on QoE estimation models, then playback quality can be improved, but the optimization may not necessarily contribute to increasing the profit of a video distribution service
Solution Approach 1:
The patent changes the optimization target from QoE metrics to KPI metrics that directly reflect business success including profit, user satisfaction, and service performance. This alignment ensures that parameter optimization directly contributes to business objectives rather than merely improving technical playback quality.
Solution Approach 2:
The patent replaces the subjective QoE estimation mechanism with an objective KPI-based evaluation system that uses actual distribution data. This substitution ensures that optimization decisions are based on measurable business outcomes rather than theoretical quality estimates.
3Measurement precision
If a significance test is performed for each combination of moving-image parameters, then optimal parameters can be selected based on actual KPI data, but the calculation and testing process becomes complex
Solution Approach 1:
The patent segments the evaluation process into distinct steps: data collection, data storage, average calculation for each parameter combination, significance testing, and final selection. This segmentation makes the complex process more manageable and systematic, allowing each step to be optimized independently.
Solution Approach 2:
The patent performs preliminary data collection and storage of KPI values for multiple parameter combinations before conducting the significance test. This preliminary action organizes the data in advance, making the subsequent statistical analysis more efficient and reducing the overall computational complexity.
Data Source
AI summary
A moving-image parameter selection device includes a storage unit configured, for each distribution of a video in a past, to associate a combination of values of moving-image parameters used for the distribution, with a KPI related to the distribution, to store the associated combination; and a selector configured, for each of the combinations, to calculate an average of the KPI, to execute a significance test for each of the averages of the KPIs with respect to one of the averages of the KPIs of the combinations, and based on results of the significance tests, to select a part of combinations among the combinations, and thereby, enables selection of moving-image parameters that contribute to increasing the profit of a video distribution service.