Design And Implementation Of Collaborative Filtering Recommendation System Based On Hypergraph Neural Network

With the development of the Internet,it provides users with more and more diverse services,and the amount of data generated is also increasing.In the massive amount of information,users cannot obtain the required information in a timely and effective manner.On the one hand,users will be at a loss when faced with a large amount of information.On the other hand,the service provider cannot get in touch with users in time,and cannot establish a good service and served relationship with users.In this context,the recommendation system came into being.It recommends items that meet the user’s interests and hobbies based on the user’s historical behavior,and thus is favored by the majority of platforms.At present,the recommendation based on collaborative filtering needs to calculate the collaborative information between users based on a large amount of existing data.With the further deepening of the research on recommendation algorithms,the existing collaborative filtering recommendation based on graph neural network encounters serious data sparse problems and the ability to model complex relationships is also deteriorating.Based on this,this paper studies a two-channel collaborative filtering recommendation algorithm based on hypergraph convolution and a user multi-behavior self-supervised recommendation algorithm based on attention mechanism,and explores the above existing problems.The specific content and results are as follows:(1)A dual-channel recommendation algorithm model based on hypergraph neural network is proposed.Compared with the existing recommendation algorithm based on hypergraph convolution,this algorithm constructs two different channels of graph convolution and hypergraph convolution,which can effectively learn user groups The common characteristics and individual characteristics of interest also strengthen the learning of the diversity characteristics of items.Finally,the effectiveness of the method is verified by experiments.(2)A multi-behavior self-supervised recommendation algorithm model based on attention mechanism is proposed.Based on the network structure of hypergraph,the model combines hypergraph attention and selfsupervised learning to make the model focus on the beneficial behavioral semantic information in hyperedge.Compared with previous graph-based recommendation algorithms,in this method,nodes and hyperedges,and features between hyperedges and graphs are all assigned different weights according to attention,which can make more effective use of behavioral information and achieve better recommendation results.(3)Implemented a real-time recommendation prototype system.This paper develops the front-end page and back-end system based on the Web,providing users with functions such as movie browsing,rating and personalized recommendation.At the same time,the system also integrates many different recommendation algorithm models,aiming to provide users with better recommendation services.