Content Recommendation System Using Third-Party Interaction Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional content recommendation systems rely heavily on users' prior history, which can lead to suboptimal results when there is no or limited prior interaction data, and may exclude content that users would be interested in, especially in areas like music and movies where usage data is weak due to factors like bootlegging and low digital sales.
Innovation Solution
A system that analyzes user interactions from other services to generate personalized content recommendations by mapping interactions with content from secondary service servers to available content on a content management server, using techniques like title parsing, metadata matching, and content fingerprinting, while ensuring user privacy and control over their data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional recommendation systems rely heavily on users' prior history, then they can provide personalized recommendations, but they produce suboptimal results when there is no or limited prior interaction data
Solution Approach 1:
The patent introduces interaction data from third-party services as an intermediary source to bridge the gap when direct user history with the content service is limited. By mapping interactions from other services (e.g., viewing behavior from social media platforms) to the target service's content catalog, the system generates recommendations even without extensive direct user history, thereby improving adaptability while maintaining recommendation quality
Solution Approach 2:
The system implements multi-functionality by utilizing interaction data from multiple different service types (social media, video platforms, music services) for the same recommendation purpose. This universal approach allows the system to aggregate diverse interaction signals across different platforms and map them to a unified content recommendation framework, improving performance across various data availability scenarios
2Reliability
If conventional systems focus on direct service usage data, then they maintain data accuracy, but they exclude content that users would be interested in due to weak usage data from factors like bootlegging and low digital sales
Solution Approach 1:
Third-party service interaction data acts as an intermediary that captures user interests not reflected in direct service usage. For example, if a user views content on a social media platform or video site before purchasing or engaging with it on the target service, this intermediary data reveals latent interests that would otherwise be lost, allowing the system to recommend content beyond what direct usage statistics show
Solution Approach 2:
The system performs preliminary mapping and analysis of third-party interaction data before generating recommendations. By pre-processing and correlating interaction patterns from multiple services with the target content catalog in advance, the system prepares enriched user profiles that capture interests even before direct service interactions occur, preventing loss of user interest information
Data Source
AI summary
A server system, which manages distribution or download of content, may obtain data relating to interactions between a user and one or more other server systems providing services that are different from services provided by the server system. The server system may then analyze the obtained interactions related data, with the analysis comprising identifying content accessed, obtained, or used by the user during the interactions between the user and the one or more other server systems. The server system may then map that content to one or more other contents available in the server system, and may generate, based on that mapping, recommendation information personalized for the user.