Research And Application Of Hybrid Recommendation Technology In E-commerce Platform Of Edible Agricultural Products

E-commerce based on agricultural products plays an important role in rural revitalization.How to recommend agricultural products that users are interested in has become a current research hotspot.The effect of traditional recommendation technology in complex environment is not ideal.The hybrid recommendation technology is to mix a variety of recommendation technology in a certain way,which can better deal with the increasingly complex recommendation requirements.To meet the actual needs of agricultural product recommendation,this paper proposes HACF(Hybrid recommendation based on Apriori and Collaborative Filtering technology)recommendation model based on hybrid recommendation technology.The basic model of HACF is mixed by parallel weighting,and the seasonal function is proposed to strengthen the influence of seasonal on agricultural product recommendation.The main research contents and achievements of this paper are as follows:(1)Research and analyze the theories of common recommendation technology.Focuses on the research of Apriori recommendation technology,recommendation technology based on collaborative filtering and hybrid recommendation technology,summarizes the relevant research status,and lists some representative research results.(2)Carry out the research and analysis of the basic recommendation model.After filtering and preprocessing the agricultural product transaction data set,Apriori recommendation technology,User-CF recommendation technology and Item-CF recommendation technology are introduced in detail and used to construct and train recommendation models respectively.The trained model will be used as the basic model of the subsequent hybrid recommendation model.(3)Carry out research on agricultural product recommendation model based on hybrid recommendation technology to meet the demand of agricultural product recommendation.Firstly,considering the seasonal characteristics of agricultural products trading,combined with the traditional mathematical model,the seasonal function is constructed to enhance the influence of season on agricultural products recommendation in the recommendation system.Secondly,the construction and training of HACF hybrid recommendation model are carried out.In this paper,the output of the basic recommendation model and the weight of the model are used as the input of the HACF hybrid recommendation model,and the seasonal function is used as the disturbance function to construct and train the model.The experimental results show that the HACF model based on the hybrid recommendation technology is effective.(4)Carry out the research and development of the e-commerce platform for edible agricultural products.Based on HACF hybrid recommendation model,this paper designs and implements the e-commerce platform for edible agricultural products based on hybrid recommendation technology and deploys the service to the cloud server.The platform based on hybrid recommendation technology consists of 9functional modules including We Chat,PDA and Web,which provide services for client users and management users respectively.The e-commerce platform developed can better meet the needs of users in the agricultural product market.