Design And Implementation Of Agricultural Product Recommendation System

With the development of Internet technology,online shopping slowly replaces traditional shopping,and there are more and more e-commerce platforms for different users and different products.Users can easily and quickly search various e-commerce platforms for the required product information and services.However,with the increase of users and products,these e-commerce platforms face the problem of information overload.How to quickly find the required information in the massive data and how to quickly find the information of interest to users in the massive information is the biggest problem in front of the major platforms.In this case,the personalized recommendation system was born.Personalized recommendation system is a basis Information users fill in at the time of registration,preferences for goods,purchase records and collection records,targeted to provide the product information they may need a technology.The use of recommendation system not only provides users with more interesting projects,but also brings a lot of benefits to Internet merchants.Through the analysis of e-commerce service platform and e-commerce trading platform in China,it is found that the online trading market of agricultural products in China has formed a certain scale,and has made remarkable achievements in the construction of e-commerce platform for agricultural products.Although China already has a large number of websites selling agricultural products,there are not many of these websites that can make personalized recommendations for users and serve users,in other words,the recommendation system for agricultural products needs to be improved.Through the agricultural product recommendation system,merchants can win in the fierce competition in the agricultural product market and create great commercial value.On the other hand,agricultural recommendation systems can reduce search the time required for the goods,which facilitates the purchase of users.Therefore,it is necessary to develop an agricultural product recommendation system which is convenient for agricultural products trading.This paper introduces the development status and background of recommendation system in detail,and also introduces the kinds of recommendation system widely used in decades.The basic architecture of Hadoop platform and the implementation of recommendation algorithm in Mahout framework are discussed.Through the distributed management of agricultural product data set through the Hadoop platform,the recommended results are obtained by using the recommendation algorithm in the Mahout framework to calculate the data.Using the data migration tool Sqoop as a bridge,the results are automatically migrated to the My SQL database,and feedback is timely when the user needs it.The main results of this paper are as follows:(1)To solve the problem of user cold start in the agricultural product recommendation system by collecting the personal information that users fill in at the time of registration,and to recommend it through the original data of new users.(2)The actual demand of the purchasers of agricultural products is analyzed,and the corresponding recommendation algorithm is selected according to the characteristics of different demand scenarios and recommendation algorithms.(3)Analyzing the whole structure of agricultural product recommendation system and constructing a system based on Hadoop platform to recommend agricultural products by Mahout collaborative filtering algorithm.Taking Hadoop as the platform of software development,the bridge between HDFS and My SQL database is constructed by Sqoop,and the recommended results are generated by using the Mahout machine learning framework,and the results are read into the database for use.