Ad Selection via Session Attributes in Video Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video streaming systems face challenges in recommending content to users without user identity information, limiting the effectiveness of ad selection and video recommendation, especially in anonymous sessions where user preferences are unknown.
Innovation Solution
A method for determining video programs and ad selection based on attributes of the current streaming session, excluding user identity parameters, by analyzing session data to automatically select video advertisements and subsequent video programs, such as those based on popularity or subscription status, and initiating new streaming sessions without requiring further user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user identity information is collected and used for ad selection and video recommendation, then recommendation effectiveness and ad targeting precision are improved, but user privacy protection and system accessibility for anonymous users deteriorate
Solution Approach 1:
The system segments user information into two categories: identity-based parameters (user profiles, historical data) and session-based parameters (current streaming session attributes). For anonymous users, only session-based parameters are utilized, allowing the system to provide targeted recommendations without requiring personal identity information. This segmentation enables the system to maintain ad targeting precision while supporting anonymous users.
Solution Approach 2:
The patent introduces session attributes as an intermediary between anonymous users and the recommendation system. Instead of directly using user identity information, the system uses session-based attributes (such as currently playing video, streaming duration, device type) as mediators to infer user preferences and make recommendations. This intermediary approach enables effective ad selection without compromising user privacy or requiring authentication.
2Measurement precision
If user identity parameters are required for content recommendation, then personalization accuracy is improved, but ease of access and user convenience deteriorate
Solution Approach 1:
The system enables anonymous users to receive personalized recommendations through self-service mechanisms. By analyzing session attributes automatically without requiring user input or authentication, the system provides personalized content recommendations as a self-service function. Users simply need to start streaming, and the system automatically generates recommendations based on their current session behavior, eliminating the need for manual profile setup or authentication.
3Adaptability or versatility
If anonymous sessions are supported without user identification, then system accessibility and audience reach are improved, but ad selection effectiveness and content recommendation precision deteriorate
Solution Approach 1:
The system employs dynamic ad selection and content recommendation based on real-time session attributes rather than static user profiles. For anonymous users, the system continuously monitors and analyzes current session dynamics (such as video playback progress, device characteristics, streaming quality) to adaptively generate recommendations and select advertisements. This dynamic approach compensates for the lack of historical user data by leveraging real-time behavioral signals.
Solution Approach 2:
The patent changes the parameters used for recommendation from identity-based parameters (user demographics, historical preferences) to session-based parameters (current video attributes, streaming session characteristics, device properties). This parameter substitution allows the system to maintain recommendation precision for anonymous users by focusing on attributes that are available and relevant during the current streaming session, thereby supporting broad system accessibility without sacrificing ad selection effectiveness.
Data Source
AI summary
In a streaming video system, a method determining a sequence of video programs playing on an interactive player device may include streaming a first video program including an ad break from a computer server to an interactive player device during a first streaming session. The method may further include automatically selecting a video advertisement or next program from a set of prospective video advertisements for play during the ad break or available programs, based at least in part on the first video program exclusive of any parameter based on user identity. The method may include providing the video advertisement to the interactive player device in the first streaming session at the ad break. Bases for selecting the next ad or program may include program popularity, rate of change in popularity, or other parameter. The method may be embodied in a computing apparatus or system.