Game Stream Recommendation System for Spectator Engagement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spectators of video games often face challenges in finding interesting gameplay sessions in real-time, as they may join a session too late or miss more engaging activities happening in other streams, and the increasing number of available streams makes it difficult to find compelling content to watch.
Innovation Solution
A system and method that utilize machine learning to analyze gameplay streams, providing personalized recommendations to spectators based on their preferences and predicting future interesting activity, allowing them to spectate at opportune times, and offering a helper mode for assisting players in gameplay.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If spectators join game streams randomly or without guidance, then they can start watching immediately, but they may miss interesting gameplay activity or spend time watching uninteresting content
Solution Approach 1:
The system performs preliminary analysis of gameplay streams using machine learning models to predict future interesting activity before spectators make their selection. This allows the system to pre-process and evaluate multiple streams, then present only those predicted to be interesting, saving spectators time and eliminating the need to manually search through uninteresting content
Solution Approach 2:
The patent introduces a recommendation system as an intermediary between spectators and game streams. This intermediary analyzes stream characteristics, predicts interesting activity, and filters streams based on spectator preferences, thereby mediating the connection between spectators and content without requiring spectators to manually evaluate each stream
2Quantity of substance
If the number of available video game streams increases, then more content is available for spectators, but it becomes increasingly difficult to find interesting gameplay to observe
Solution Approach 1:
The system implements feedback mechanisms where spectator selections and engagement patterns are continuously analyzed to refine machine learning models. This feedback loop allows the recommendation system to improve its accuracy over time in predicting which streams will be interesting to specific spectators, making the process of finding interesting content easier as the system learns from accumulated data
Solution Approach 2:
The patent replaces the manual mechanical process of spectators browsing and evaluating streams with an automated machine learning-based recommendation system. This substitution uses computational models to analyze stream characteristics and predict interesting activity, eliminating the need for spectators to manually search through increasing numbers of streams
Data Source
AI summary
In some implementations, a method is provided, performed by at least one server computer, for providing spectating of gameplay of a video game, including: receiving a plurality of game streams, each game stream being generated from an executing session of a video game; analyzing each game stream, to recognize gameplay activity depicted in each game stream; obtaining spectating preferences of a user; determining a prioritization of the plurality of game streams based on the recognized gameplay activity and the spectating preferences of the user; providing, over a network via a client device operated by the user, a recommendation of at least some of the game streams for spectating based on the determined prioritization; responsive to receiving, over the network from the client device, a selection of one of the recommended game streams, then providing, over the network to the client device, the selected game stream for spectating by the user.