Personalized Ad Display Using a Terminal-Side Knowledge Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing advertisement placement methods fail to consider individual user preferences, leading to suboptimal recommendation homogeneity and reduced economic benefits for providers.
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 content recommendation, ensuring user privacy is protected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If advertisement placement uses group profiling based on service data, then advertisement coverage can be achieved, but individual user preferences are not considered leading to recommendation homogeneity
Solution Approach 1:
The patent segments the advertisement recommendation process into two parts: group-level profiling done by the server and individual-level re-ranking done by the terminal. This segmentation allows individual preferences to be considered while keeping sensitive personal data processing localized, thus resolving the contradiction between adaptability and privacy protection.
Solution Approach 2:
The patent introduces a personal knowledge graph as an intermediary data structure that stores user preferences locally on the terminal. This intermediary enables individualized recommendation without requiring raw personal data to be transmitted to the server, thus achieving both adaptability and privacy protection.
2Measurement precision
If terminal side processes user data for personalization, then advertisement placement accuracy improves, but user privacy security may be compromised
Solution Approach 1:
The patent implements self-service by enabling the terminal to autonomously perform re-ranking of advertisements using locally stored user preference data. This eliminates the need to transmit sensitive personal information to the server while still achieving personalized recommendation, thus improving both accuracy and privacy security simultaneously.
3Productivity
If server screens advertisements based on group profiling, then advertisement delivery efficiency is maintained, but recommendation homogeneity reduces economic benefits
Solution Approach 1:
The patent applies preliminary action by pre-building personal knowledge graphs on terminals that capture individual user characteristics. This preliminary local processing preserves individual traits without requiring real-time server analysis, thus maintaining delivery efficiency while preventing recommendation homogeneity.
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).