Video Delivery Problem Ranking via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High traffic Internet-based media streaming systems face challenges in discovering the causes of re-buffering events and mitigating them due to the sheer quantity of data and numerous variables involved, leading to volatile performance and degraded playback quality for users.
Innovation Solution
An automated machine learning-based approach is employed to identify and evaluate problems associated with re-buffering by gathering large amounts of data from both server and client ends, processing it to generate a table that relates quality information to re-buffer events, and using machine learning techniques to determine weighted values for features contributing to re-buffering, allowing for ranking and suggesting changes to improve playback quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional manual analysis methods are used to identify re-buffering causes, then analysis precision can be maintained, but productivity is severely limited due to the sheer quantity of data and numerous variables
Solution Approach 1:
The patent replaces manual mechanical analysis with automated machine learning algorithms. The system collects playback information including re-buffer events and network conditions, then uses automated ML models to identify causal factors, eliminating the need for manual data sifting while handling the sheer quantity of data and numerous variables efficiently
Solution Approach 2:
The patent introduces an intermediary analysis system that sits between raw playback data and actionable insights. This intermediary layer collects data from multiple sources (playback information, network conditions, device characteristics), processes it through ML algorithms, and outputs ranked causal factors, simplifying the complexity for end users
2Measurement precision
If comprehensive data collection from server and client ends is implemented, then measurement precision of playback quality issues is improved, but device complexity and data processing requirements increase
Solution Approach 1:
The patent segments the data collection system into distinct components: server-end data collection, client-end data collection, and centralized analysis. Playback information is divided into specific categories (re-buffer events, network conditions, device characteristics), allowing comprehensive data gathering while managing complexity through modular organization
Solution Approach 2:
The patent creates a universal data collection framework that operates across server and client ends with consistent protocols. The same analytical model processes data regardless of source, and the system handles multiple types of playback information through a unified approach, reducing overall system complexity
3Productivity
If machine learning techniques are used to analyze playback information, then productivity in identifying re-buffering causes is improved, but manufacturing precision of analysis results may be affected by model accuracy
Solution Approach 1:
The patent implements feedback mechanisms where analysis results are validated against actual playback outcomes. The system continuously refines its ML models based on accumulated data, comparing predicted causal factors with actual re-buffering patterns to improve accuracy over time while maintaining high processing speed
Solution Approach 2:
The patent performs preliminary data processing and feature extraction before applying ML analysis. By pre-processing playback information to extract relevant features and organize data into standardized formats, the system prepares high-quality input for ML models, improving both processing efficiency and analysis accuracy
Data Source
AI summary
Systems and methods for determining video infrastructure delivery problems using machine learning are presented. In an aspect, a system includes a reception component configured to receive information regarding videos streamed by the system to devices, wherein the information identifies video playback events at the devices and re-buffer events respectively associated with the video playback events. The system further includes a quality component configured to identify features related to quality of the playback events at the devices based on the information, and an analysis component configured to determine probabilities of occurrence of the re-buffer events based on different combinations of the features, and determine weighted values for each of the features that reflect their contribution to the probabilities of occurrence of the re-buffer events based on the different combinations of the features.