Recommendation Factor Computation for Event Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional recommendation systems face challenges in accurately determining a user's interests due to insufficient feature extraction, leading to inaccurate and non-customized recommendations.
Innovation Solution
A novel algorithm is introduced to compute a recommendation factor by considering both the user's execution/acceptance frequency and the common level of the recommendation/event, which improves the accuracy of recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional collaborative filtering is used to determine recommendations based on execution frequency, then the recommendation system is simple to implement, but the accuracy of recommendations deteriorates because it cannot distinguish between commonly executed events and user-specific interests
Solution Approach 1:
The patent segments the execution frequency metric into two distinct components: user-specific execution frequency (how often a specific user executes an event) and common execution frequency (how often events are executed across all users). This segmentation allows the system to differentiate between popular events and user-specific interests, thereby improving recommendation accuracy without requiring overly complex feature extraction mechanisms.
Solution Approach 2:
The patent introduces a new dimension to the recommendation space by computing a recommendation factor that combines user-specific execution frequency with inverse common execution frequency. This dimensional transformation converts the single-dimensional collaborative filtering approach into a two-dimensional analysis, enabling the system to identify events that are both frequently executed by a user and relatively uncommon across the user base, thus improving recommendation precision.
2Measurement precision
If execution frequency alone is used to determine recommendations, then the calculation is simple, but the system cannot distinguish whether a user executed an event due to interest or because it was a commonly executed event
Solution Approach 1:
The patent segments the execution frequency metric into two distinct components: user-specific execution frequency (how often a specific user executes an event) and common execution frequency (how often events are executed across all users). This segmentation allows the system to differentiate between popular events and user-specific interests, thereby improving recommendation accuracy without requiring overly complex feature extraction mechanisms.
Solution Approach 2:
The patent introduces a new dimension to the recommendation space by computing a recommendation factor that combines user-specific execution frequency with inverse common execution frequency. This dimensional transformation converts the single-dimensional collaborative filtering approach into a two-dimensional analysis, enabling the system to identify events that are both frequently executed by a user and relatively uncommon across the user base, thus improving recommendation precision.
Data Source
AI summary
A method, performed by at least one processor of a computing device, for recommending an event, the method including determining a frequency of execution of the event by a user among a plurality of users; determining an inverse user frequency of execution of the event based on the plurality of users who have executed the event; generating a recommendation factor by modifying the frequency of execution of the event by the user based on the inverse user frequency of execution of the event; and outputting a recommendation for the event based on the recommendation factor.