Design And Implementation Of Agricultural Product E-commerce Platform Based On Recommendation Algorithm

The SARS epidemic in 2003 has led many people to abandon the traditional shopping mode,stay indoors,far away from the bustling supermarkets and shopping malls,and e-commerce,represented by Taobao,has risen strongly.With the birth of online payment means,e-commerce has gradually penetrated into all aspects of people's lives,and e-commerce has also started from B2 B C2B,C2 C,o2o,B2B2 C and other business models are gradually born.With the deepening of "Internet plus agriculture",more and more agricultural products are moving towards the Internet through e-commerce and the public.This greatly shortens the distance between consumers and agricultural production lines,not only facilitates the sale of agricultural products,but also improves the quality of consumers' quality.The purpose of recommendation system is to help people quickly and accurately select their own target products,and recommend products that have not been touched due to the user's life domain to users.The traditional recommendation algorithms of e-commerce mainly include "user based collaborative filtering algorithm" and "commodity based collaborative filtering recommendation algorithm",which can achieve the main goal of recommendation to a certain extent.With the increasing variety,quantity and scale of agricultural products in the network,the conventional recommendation of agricultural products e-commerce has been unable to meet the demand,and a new recommendation idea is urgently needed to solve the recommendation problem.Based on the traditional "user based collaborative recommendation algorithm","commodity based collaborative filtering recommendation algorithm" and convolutional neural network,this paper adopts the design of J2 EE and the design mode of MVC,combined with the SSM framework,to realize the application development of the collaborative filtering recommendation e-commerce system based on convolutional neural network.This paper includes the following aspects of work.This paper proposes a collaborative recommendation filter system based on convolutional neural network.Compared with the traditional collaborative filtering,the idea of this paper is to adopt different recommendation methods in the adaptive scene.In the "guess what I like" similar "based on user recommendation" scenario,by collecting the historical preferences of the target users,combining with the product product relationship extracted by the convolution neural network,the similarity between user feature preferences and product features is calculated,and finally topk is generated.In the scene of "based on item recommendation" similar to "recommended and designated goods",by collecting the parameters of the target goods,analyzing the characteristics,combining the relationship between the goods and the goods extracted by convolution neural network,the similarity between the target goods and the designated goods is calculated,and finally the top_k is generated.According to the design mode of MVC and the SSM framework,the e-commerce trading platform for agricultural products is developed and the final test is completed.