Terminal-Side Knowledge Graphs for Personalized Advertisement Display
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing advertisement placement methods fail to consider individual user preferences, leading to suboptimal advertisement placement effects and potential privacy issues due to the lack of personalized data processing.
Innovation Solution
An advertisement display method that constructs a personal knowledge graph on the terminal side using user data to enhance advertisement selection, incorporating a re-ranking model for personalized advertisement recommendations while protecting user privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If advertisement placement is based on group profiling data from servers, then advertisement coverage and delivery efficiency are improved, but individual user preferences are not considered leading to suboptimal placement effects
Solution Approach 1:
The system segments advertisement placement into two levels: group-level profiling performed by the server for efficiency, and individual-level refinement performed by the terminal device for precision. The terminal device divides the group-profiled advertisement list into individualized recommendations based on personal knowledge graphs, resolving the contradiction between group-level efficiency and individual-level accuracy.
Solution Approach 2:
The patent adds a new dimension of personalization by constructing knowledge graphs that capture temporal patterns and contextual relationships of user behavior. This transforms the flat group-profiling approach into a multi-dimensional individualized recommendation system that considers time, context, and user preferences simultaneously.
2Measurement precision
If individual user data is collected and processed for personalized recommendations, then advertisement placement accuracy is improved, but user privacy security may be compromised
Solution Approach 1:
The patent extracts only the necessary minimal personal data required for knowledge graph construction on the terminal device, rather than collecting comprehensive user data on servers. By performing data processing locally and extracting only essential patterns for advertisement selection, the system achieves personalization while minimizing privacy exposure.
Solution Approach 2:
The terminal device performs self-service by constructing and maintaining its own knowledge graph locally without requiring user data to be uploaded to external servers. The device independently processes its own usage data, generates personalization patterns, and makes advertisement selection decisions, thereby eliminating the need to transmit sensitive personal information externally.
3Object-affected harmful factors
If comprehensive personal data is processed on the terminal device, then user privacy is protected, but data processing complexity and resource consumption increase
Solution Approach 1:
The system performs partial data processing on the terminal device by focusing only on constructing knowledge graphs for advertisement-related patterns rather than processing all user data comprehensively. This selective approach achieves privacy protection through local processing while avoiding the complexity burden of complete data analysis.
Data Source
AI summary
The method includes: An electronic device 100 obtains first personal data (S1001); the electronic device 100 constructs a personal knowledge graph based on the first personal data (S1002); the electronic device 100 obtains parameter information of first advertisement content from an advertisement server 200 (S1003); the electronic device 100 obtains parameter information of second advertisement content from the parameter information of the first advertisement content based on the personal knowledge graph (S1004); the electronic device 100 obtains the second advertisement content based on the parameter information of the second advertisement content (S1005); and the electronic device 100 displays the second advertisement content on a display (S1006).