On-Device Content Generation for Secure Context-Aware Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for selecting and providing third-party content items on client devices lead to data security risks and unnecessary network and computing resource utilization due to frequent transmission of device context information to content providers.
Innovation Solution
Client devices generate and select on-device content items using local context information, reducing the need for external data transmission and enhancing security by only transmitting interaction data to content providers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If device context information is transmitted frequently to third-party content providers to provide relevant content items, then content relevance is improved, but network resource utilization increases and data security risks worsen
Solution Approach 1:
The system performs preliminary action by generating and caching multiple on-device content items in advance based on historical context information. When a content request occurs, the pre-generated content items are immediately available for selection without requiring real-time network communication, thus reducing network resource utilization while maintaining content relevance.
Solution Approach 2:
The patent introduces an intermediary mechanism by using on-device content generation models that act as mediators between the device context information and the content items. The model processes context information locally and generates content items without requiring direct communication with third-party content providers, reducing network traffic while preserving content relevance.
2Adaptability or versatility
If device context information is transmitted to third-party content providers for content selection, then content personalization is improved, but data security risks worsen
Solution Approach 1:
The system extracts the content generation function from the third-party content provider and places it locally on the device. By taking out the context information processing from the external server and performing it locally through on-device models, the patent eliminates the need to transmit sensitive device context information to external parties, thus reducing data security risks while maintaining content personalization capabilities.
Solution Approach 2:
The device performs self-service by using locally stored content generation models to create personalized content items based on its own context information. The device serves its own content needs without requiring external assistance, thereby keeping sensitive information local and reducing exposure to security risks associated with data transmission to third parties.
3Loss of energy
If on-device content generation is implemented, then network resource utilization decreases and security improves, but device computational load increases
Solution Approach 1:
The system applies partial action by generating only the necessary number of content items (e.g., a small set of candidate items) rather than exhaustively generating all possible content. This selective generation approach reduces the computational burden on the device while still providing sufficient content options for selection, thereby balancing local processing requirements with network resource savings.
4Adaptability or versatility
If multiple network transmissions occur to provide content items (context information transmission and content delivery), then content relevance is improved, but the number of network transactions increases
Solution Approach 1:
The patent merges the context information transmission and content delivery processes into a single local operation. By combining the functions of context analysis and content generation into an on-device model, the system eliminates the need for separate network transmissions for context submission and content retrieval, reducing the total number of network transactions while maintaining content relevance.
Data Source
AI summary
At least one aspect of the present disclosure is directed to systems and methods of generating on-device content items to improve security and network utilization. A client device can access content resources and category information from a data structure in the memory the memory of the client device. The client device can generate on-device content items based on the content resources and the category information. The client device can receive an indication to display a content item in an application executing on the client device. The client device can scan the client device for context information to create a relevant device context. The client device can select an on-device content item based on the relevant device context. The client device can provide the selected on-device item to the application for display, detect an interaction with the displayed content item, and update a content selection model or a content generation model.


