A product recommendation method based on user portrait

Abstract

The invention discloses a method for recommending commodities based on user portraits, which includes: acquiring commodity information and Internet celebrity information through big data collection; clustering the commodity information and user information through the Internet celebrity information, and manually Combined with intelligent algorithms to generate user portraits and product features; the user information is the information of Internet celebrity fans; the user portraits and product features are recorded, and by continuously collecting and clustering the product information and user information, Increase the number of samples and expand the number of user portraits and product features; match the coincidence degree of the user portrait and the product features through a weight algorithm, establish and save the relationship topology between the user portrait and the product features; According to the user portrait, the product corresponding to the product information is located according to the relationship topology; after manual review, it is pushed to Internet celebrities or merchants for two-way screening.

Description

technical field [0001] The invention relates to the field of e-commerce development, in particular to a product recommendation method based on user portraits. Background technique [0002] With the prevalence of Internet celebrity culture and the rise of the live delivery industry, a large number of traditional merchants and self-media Internet celebrities have entered this industry. However, different Internet celebrities have different user groups, and different merchants have different target customers. Every time a product is matched with an Internet celebrity and finally broadcast, it is a trial-and-error behavior with huge risks. Internet celebrities may Lose popularity due to inappropriate product selection, and businesses may also suffer losses due to inappropriate cooperation. [0003] In order to solve the problem of matching Internet celebrities and products, this system will collect all major KOL Internet celebrities and product information on the entire network...

AI Technical Summary

Problems solved by technology

[0009] CN110996110A provides a commodity adjustment method and device in the live broadcast process, which relates to the technical field of live broadcast marketing, by obtaining the first purchase situation of the first commodity in the first live broadcast e-commerce platform; according to the first purchase situation, determine the target user portrait; determine the first purchase preference according to the target user portrait; determine the second commodity according to the first purchase preference; determine whether the second commodity satisfies the first preset condition; when the first preset condition is not met When setting the conditions, the second commodity is listed in the first live broadcast catalog, thereby solving the problem that in the prior art, live broadcast commodities are generally pushed and selected fixedly, and the commodity strategy cannot be adjusted in real time according to the situation of live broadcast sales, thereby The technical problems that affect the user's live shopping experience have achieved the ability to adjust the product strategy in real time according to the live sales situation, greatly improving the user's desire to buy and the technical effect of the live shopping experience

[0010] In the above method, you can only look for competing products for products that have already been sold by Internet celebrities and have good performance. On the one hand, under a single category and niche brands, there are few competing products to choose from. Seeking more cross-border attempts cannot provide a solution, and merchants can only always screen cooperative Internet celebrities within a fixed range of similar purposes, which has limitations

Images