Real-Time Recommendation Switching Using Gaze and Attention Levels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information recommendation systems fail to switch recommended information in real time without explicit user operations, leading to suboptimal user engagement in physical stores.
Innovation Solution
An information search support system that utilizes user information acquisition, interest level calculation, attention level calculation, and recommendation suitability level calculation to determine the timing of switching recommended information on display terminals, based on line of sight and behavior analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If information is switched only when explicit user operations are performed, then user convenience is maintained, but real-time switching of recommended information cannot be achieved
Solution Approach 1:
The system automatically switches recommended information by detecting user line of sight and calculating interest/attention levels without requiring explicit user operations. The display information control unit autonomously determines switching timing based on calculated metrics, enabling the system to serve itself rather than waiting for user commands.
Solution Approach 2:
The system continuously monitors user line of sight information and calculates interest levels and attention levels as feedback metrics. This feedback loop enables real-time adjustment of recommended information based on actual user engagement, allowing dynamic switching without explicit user input while maintaining responsiveness to user needs.
2Productivity
If recommended information is switched in real-time based on user behavior, then user engagement is enhanced, but system complexity increases
Solution Approach 1:
The system divides the complex task of real-time recommendation into distinct functional modules: line of sight information acquisition, interest level calculation, attention level calculation, suitability level calculation, and display control. Each unit handles a specific aspect of the process, making the overall system more manageable and maintainable despite its real-time capabilities.
Solution Approach 2:
The display information control unit serves multiple functions: it manages display content, determines switching timing based on calculated metrics, and coordinates with the recommendation system. This multi-functionality reduces the need for separate dedicated components, thereby managing system complexity while achieving real-time adaptive recommendations.
3Measurement precision
If multiple calculation units are used to estimate user interest and attention levels, then recommendation accuracy is improved, but processing time increases
Solution Approach 1:
The system performs preliminary calculations by continuously estimating user interest levels based on line of sight information before actual switching decisions are made. This ongoing preliminary assessment allows the system to have readiness data prepared, reducing the processing time required at the moment of switching while maintaining high recommendation accuracy.
Solution Approach 2:
The calculation units operate continuously rather than intermittently, maintaining constant estimation of interest and attention levels. This continuous operation ensures that accurate recommendation data is always available, eliminating delays that would occur from periodic recalculations, thereby achieving both high accuracy and responsive switching.
Data Source
AI summary
Information search support system which estimates in real time a user interest level which indicates a level of interest of the user in each piece of recommended information based on user information acquired over a predetermined time period, estimates in real time a level of attention which indicates a level to which the user is currently paying attention to each piece of recommended: information based on currently acquired user information, calculates a level of suitability for each piece of recommended information based on the interest level and the attention level, acquires new recommended information from the information recommendation system based on the user information and the interest level, and determines a timing of switching to suitable recommended information based on the suitability level, the interest level, and the new recommended information, and requests the information display terminal to switch the recommended information being displayed.