Product recommendation method and device based on multi-model fusion and medium

Abstract

The invention discloses a product recommendation method and device based on multi-model fusion and a medium, and the method comprises the steps: determining a user interest vector of a contained first product, and obtaining product standardization feature vectors corresponding to all products; obtaining a first product recommendation table for the user according to the similarity between the user interest vector and the product standardization feature vector; according to the product browsing data, determining product browsing times, and according to the user application product data, determining product application times of a second product of the user; obtaining a user-product matrix, performing non-negative matrix factorization training according to the user-product matrix to obtain an NMF model, and obtaining a second product recommendation table according to the NMF model; according to the product browsing data, training a product recommendation model in a gradient lifting mode; and fusing the first product recommendation table and the second product recommendation table to obtain a third product recommendation table, and performing prediction through a product recommendation model to obtain a final recommendation result.

Description

technical field [0001] The present application relates to the field of artificial intelligence, and in particular to a method, device and medium for product recommendation based on multi-model fusion. Background technique [0002] With the development of information technology and big data technology, people have entered the era of information overload from the era of information scarcity. It's getting harder and harder to visualize it to users who are interested in it. The task of the recommender system is to connect users and information, as well as various associations between users and information, and recommend appropriate information for users. [0003] The recommendation system finds the user's personalized needs by mining and analyzing the user's behavior information, so as to accurately recommend the specific item information to the appropriate user, and help the user to find the items that they are interested in but difficult to find. Various algorithms for recom...

AI Technical Summary

Problems solved by technology

Since the development of recommendation system technology, many recommendation algorithms have been proposed and used in the industry. After a lot of practice, it is difficult to have a solution that can be widely adapted to any application scenario. Each recommendation method has its limitations. For example, content-based recommendation The features extracted by the algorithm must not only ensure accuracy but also have certain practical significance, otherwise it is difficult to ensure the relevance of the recommendation results; the recommendation algorithm based on association rules has a large amount of calculation, and due to the use of user data, there is inevitably a cold start and sparsity problems; recommendation algorithms based on popularity cannot provide personalized recommendations for users

Images