Viewing Abandonment Factor Estimation Using Segmented Quality and Content Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to effectively model and estimate the causal relationship between various feature quantities related to application quality, user operation, and content factors on viewing abandonment in adaptive bit rate video distribution, often considering either quality or content impacts but not both quantitatively.
Innovation Solution
A viewing abandonment factor estimation device that uses a model incorporating feature quantities such as application quality, user operation, and content features to classify viewing abandonment factors, employing statistical or machine learning models like classification trees, random forests, logistic regression, or support vector machines to determine whether abandonment is due to quality or content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a viewing abandonment model is built using only quality factors or only content factors, then the model is simple to construct, but the estimation precision of viewing abandonment factors is insufficient
Solution Approach 1:
The patent segments the viewing abandonment factor into two distinct components: quality factor and content factor. By dividing the complex estimation problem into separate quality-related features (playback stop, buffering, resolution) and content-related features (genre, duration, popularity), the model can process each segment independently and then combine results, improving overall estimation precision without overwhelming complexity
Solution Approach 2:
The patent merges multiple feature types (quality features, content features, and their interactions) into a unified viewing abandonment estimation model. This combination allows the system to capture both quality-induced and content-induced abandonment behaviors simultaneously, achieving high estimation precision by integrating diverse data sources into a cohesive analytical framework
2Measurement precision
If multiple feature quantities including both quality and content factors are incorporated into the model, then the estimation precision improves, but the device complexity increases
Solution Approach 1:
The patent segments the viewing abandonment factor into two distinct components: quality factor and content factor. By dividing the complex estimation problem into separate quality-related features (playback stop, buffering, resolution) and content-related features (genre, duration, popularity), the model can process each segment independently and then combine results, improving overall estimation precision without overwhelming complexity
Solution Approach 2:
The patent transforms multiple diverse feature quantities into standardized parameters suitable for machine learning processing. Quality features (playback stop count, buffering events, resolution changes) and content features (genre categories, duration segments, popularity rankings) are converted into normalized numerical parameters, enabling efficient model processing while maintaining the precision benefits of multi-factor analysis
Data Source
AI summary
A viewing abandonment factor estimation device includes a memory; and a processor configured to include an estimation model for estimating a factor, the estimation model including a plurality of feature quantities measurable for viewing of a video relevant to an adaptive bit rate video distribution as inputs and the factor of the viewing abandonment in the viewing as an output.