Cosmetic personalized product matching method and device and computer

Abstract

The invention belongs to the technical field of personalized service algorithms, and particularly relates to a cosmetic personalized product matching method and device and a computer, and the method comprises the steps: obtaining the skin basic feature information data of a user; inputting the user skin basic feature information data into a preset skin score table to obtain a user skin basic score, and storing the user skin basic score into a user skin archive score table; inputting the specified product attribute parameters into the user skin archive scoring table to obtain weight scoring results corresponding to the product attribute parameters; clustering the user data by using a clustering algorithm in data mining according to a weight scoring result corresponding to the product attribute parameter to obtain a clustered data result; and according to a clustering data result, a cosmetic product classification target is obtained by using a mixed integer programming class algorithm, and cosmetic samples are distributed according to the classification target. Products and services are individually matched for each user, so that the product satisfaction of the user is met to the greatest extent.

Description

technical field [0001] The invention belongs to the field of personalized service algorithms, and in particular relates to a cosmetic personalized product matching method, device and computer. Background technique [0002] Against the background of the vigorous development of the beauty industry, a large number of new cosmetic products are constantly pouring into the market, regardless of product form, design, and color are constantly updated. At present, cosmetics consumption mainly comes from online. Therefore, in the face of a huge cosmetics product library, users need to identify, filter, and select the products they are interested in, and this process is time-consuming and labor-intensive. At present, e-commerce platforms mostly use recommendation algorithms to recommend products for users. Common algorithms include collaborative filtering and content-based recommendation. Collaborative filtering is to predict preferences based on user behavior, which can be user-based ...

AI Technical Summary

Problems solved by technology

However, in actual operation, the types and quantities of samples that the platform can provide are often limited, and the accuracy and efficiency of sample distribution are low. For brands that need to use samples to promote, it has caused cost losses.

Images