Dynamic Inference Collaboration for Cross-Device Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing content recommendation systems provide generic, device-specific recommendations that lack personalization and intelligence, failing to dynamically adapt to user scenarios across multiple devices.
Innovation Solution
A dynamic inference collaboration mechanism that receives content metadata from a content provider, identifies scenarios and keywords, and generates personalized recommendations by relating them to stored inferences across devices, using a graph database to link devices, scenarios, and content providers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If content recommendation systems provide device-specific recommendations, then recommendations can be generated quickly and simply, but the recommendations become generic and lack personalization across multiple devices
Solution Approach 1:
The patent merges inference data from multiple devices into a unified graph database structure, combining device-specific inferences with user-specific inferences to create a comprehensive recommendation system that works across devices while maintaining personalization
Solution Approach 2:
The graph database structure serves multiple functions: storing device inferences, storing user inferences, enabling cross-device recommendations, and supporting scenario-based recommendations. This universal structure resolves the contradiction by providing adaptability across devices while maintaining manageable complexity through a single unified system
2Extent of automation
If content recommendation systems use simple device-specific logic, then the system is easy to operate, but the recommendations lack intelligent analysis and are generic
Solution Approach 1:
The system automatically performs scenario identification, keyword extraction, and inference generation without requiring manual configuration. The graph database structure enables the system to self-organize recommendations by automatically relating new inferences to existing scenarios and keywords, providing intelligent analysis while maintaining ease of operation
Solution Approach 2:
The system pre-structures the recommendation space by organizing inferences into scenarios and keywords before queries arrive. This preliminary organization of data into the graph database structure enables rapid intelligent retrieval and analysis when recommendations are requested, without adding complexity to the operation interface
3Loss of information
If content recommendation systems store comprehensive content inferences, then recommendations can be highly personalized, but the system requires complex database structures and processing
Solution Approach 1:
The patent segments comprehensive content inferences into distinct, manageable components: device inferences, user inferences, scenarios, and keywords. Each segment is stored in a specific portion of the graph database with a defined structure, allowing the system to retain comprehensive information while maintaining database simplicity through organized segmentation of data types and relationships
Data Source
AI summary
An electronic device includes at least one memory and at least one processor coupled to the at least one memory. The at least one memory is configured to store a database. The at least one processor is configured to receive content metadata from a content provider based on a query from a first device. The at least one processor is also configured to identify a scenario and a keyword associated with the content metadata. The at least one processor is further configured to generate a recommendation based at least in part on content inferences associated with the first device stored in the database through relating the scenario to previously-identified scenarios and relating the keyword to previously-identified keywords. In addition, the at least one processor is configured to provide the recommendation.