Skin Care Formulation Clustering for Ingredient Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The development of new skin care product formulations is inefficient due to the complexity of selecting and proportioning ingredients, with existing computer-aided design methods struggling to optimize formulations effectively as the number of ingredients increases, leading to difficulty in finding optimal formulations.
Innovation Solution
A method and system that utilize a skin care product formulation database to acquire and process new ingredients, cluster formulations based on ingredient correlations, calculate correlation scores, and generate new formulations using a machine learning algorithm, specifically employing Bag of Words, KNN, Doc2Vec, or tf-idf, to determine the optimal blend of ingredients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the number of ingredients in skin care product formulation is increased to improve product efficacy and functionality, then the product performance is improved, but the complexity of formulation optimization and selection becomes significantly more difficult
Solution Approach 1:
The patent segments the complex formulation optimization problem into distinct processing stages: ingredient correlation analysis, formulation clustering, candidate generation, and iterative optimization. Each stage handles specific aspects of the formulation process independently, making the overall complex task manageable through division of labor among different computational modules.
Solution Approach 2:
The patent introduces an intermediary computational system that acts as a mediator between the large number of available ingredients and the final formulation. This system uses correlation analysis and clustering algorithms to bridge the gap between raw ingredient data and optimized formulations, automatically identifying meaningful relationships without requiring manual analysis of all possible ingredient combinations.
2Extent of automation
If computer aided design methods are used to optimize ingredient proportions, then the formulation process is automated, but the complexity of collocation and proportional relation optimization increases significantly with more ingredients
Solution Approach 1:
The patent performs preliminary actions by pre-calculating ingredient correlations and pre-clustering formulations before the actual optimization process. This preliminary processing organizes the data structure in advance, creating a ready-to-use framework that simplifies subsequent automated optimization tasks and reduces the computational burden during iterative design cycles.
Solution Approach 2:
The patent creates simplified copies or representations of the complex formulation problem through clustering. Instead of directly optimizing all possible ingredient combinations, it generates clustered representation formulations that capture the essential relationships, making the optimization problem more tractable while preserving the key characteristics of the original complex system.
3Reliability
If manual iteration and formulator experience are used to determine optimal formulations, then the formulation quality can be maintained, but the development process becomes time-consuming and inefficient
Solution Approach 1:
The patent implements feedback mechanisms where the computational system continuously evaluates formulation candidates, compares them against target specifications, and uses the evaluation results to guide subsequent optimization iterations. This closed-loop feedback process automatically captures and utilizes formulator expertise, maintaining formulation quality while eliminating the time-consuming manual iteration cycle.
Solution Approach 2:
The patent enables the formulation optimization process to serve itself through automated algorithms that independently perform correlation analysis, clustering, candidate generation, and evaluation without requiring continuous human intervention. The system uses its own generated data and results to drive the optimization process forward, significantly improving development efficiency while maintaining reliability through systematic automated decision-making.
Data Source
AI summary
Disclosed is a skin care product formulation development method and system. The method comprises: acquiring names and weights of new ingredients, searching formulations containing at least one of the new ingredients in a skin care product formulation database, and obtaining a plurality of candidate formulations; deleting the new ingredients contained in each of the candidate formulations to obtain a plurality of supplement formulations; clustering the plurality of supplement formulations based on ingredient correlations to obtain a plurality of cluster groups; calculating a correlation score of each cluster group to the deleted new ingredients respectively; and determining the cluster group with the highest correlation score, and generating a new skin care product formulation based on the cluster group with the highest correlation score.