Scent Recommendation System Using User Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Human nasal preferences for scents are highly variable and fluid, making it difficult for individuals to consistently choose appropriate scents, as the same scent can be perceived differently at different times due to the dynamic nature of nasal receptors.
Innovation Solution
A scent-presentation-information output system that includes a scent-provision-information storage unit, acquisition unit, clustering unit, and output unit, which classifies users based on scent preferences and provision duration to recommend scents likely to be preferred by clustering users with similar preferences, using a combination of scent evaluation information, user identification, and feedback analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If scent recommendations are based on individual user preferences, then the system can provide personalized suggestions, but the recommendations become unreliable due to the fluid and variable nature of human nasal preferences
Solution Approach 1:
The patent merges individual user preferences with cluster-based patterns. By combining personal scent evaluation data with aggregated data from similar users, the system achieves both personalization and reliability. The recommendation is generated by integrating the user's own preferences with the preferences of users in the same cluster, thereby compensating for the variability of individual nasal characteristics through statistical aggregation.
2Loss of information
If the system collects detailed scent evaluation information from individual users, then it can generate personalized recommendations, but it requires extensive data collection and processing time
Solution Approach 1:
The system performs preliminary clustering of users based on their scent evaluation information before generating specific recommendations. By pre-organizing users into clusters with similar preferences, the system reduces the processing time required during actual recommendation generation. The clustering operation is performed in advance, so when a recommendation is needed, the system only needs to query the pre-formed clusters rather than analyzing all user data from scratch.
3Reliability
If the system uses clustering to group users with similar preferences, then it can improve recommendation reliability, but it increases system complexity
Solution Approach 1:
The patent applies local quality by creating clusters with specific, localized characteristics rather than attempting to classify users by all possible attributes. Each cluster is defined by specific scent evaluation patterns that are locally relevant to that group. This approach reduces overall system complexity by focusing classification efforts on the most important and discriminatory features of user preferences, rather than attempting to account for every possible variable.
Data Source
AI summary
A scent-presentation-information output system includes a scent-provision-information storage unit, a scent-provision-information acquisition unit, a use information reception unit, a fourth-evaluation-information generation unit, a scent-presentation-information generation unit, and a scent-presentation-information output unit. The scent-provision-information storage unit stores scent provision information identifying a scent provided to a user, scent evaluation information indicating an evaluation of the scent by the user, and user identification information. The use information reception unit receives input of a use purpose of the scent and an effect for the use purpose. The fourth-evaluation-information generation unit generates fourth evaluation information based on the use purpose and the effect to be included in the scent evaluation information. The clustering unit classifies the user into a predetermined cluster. The scent-presentation-information generation unit generates scent presentation information for presenting a scent recommended to the user. The scent-presentation-information output unit outputs the scent presentation information to an output destination.