The Research And Implementation Of Grecs E-commerce Recommender System For Agricultural Products

With the development of wisdom agriculture and e-commerce, more and more people begin to use agricultural e-commerce system for online trading, recommender system is an important part of the e-commerce system. At present, each big e-commerce platform has more perfect commodity recommendation features. However, there are few indeed experts recommender system for agricultural products. Grecs(Greenhouse Recommender System) recommender system is an important part of electronic commerce system which is based on the intelligent greenhouse system. It is mainly used to recommend vegetables and fruits which users have been interested in.In this paper, main contents and results are as follows:Firstly, this paper analyses and studies the current existing constructing technology of e-commerce recommender algorithm systems, and then build the Grecs Agricultural E-commerce System which is based on the SSH server framework technology and J2 EE development platform.Secondly,this paper study and improve the existing collaborative filtering algorithm which is based on item, build the Item CF-Time Grade collaborative filtering recommender algorithm based on item with the time factor and user ratings information. Item CF-Time Grade algorithm can provide the users good feedback and let item list update faster.Thirdly, through the analysis of expert system and combine with agricultural products category features, this paper design and build the Greenhouse-Expert agricultural experts recommender algorithm based on ontology. Greenhouse-Expert algorithm can offer more professional and accurate recommended results, and solve the cold start problem of the recommender system. This paper put forward this building methods in constructing the Greenhouse-Expert agricultural experts recommender algorithm based on ontology, this method combines the ideas of expert knowledge and recommender algorithm.Finally, mix these two algorithm, combines the advantages of these two algorithms and get the final Expert Item CF-Time Grade hybrid recommender algorithm. Then this paper realize the Expert Item CF-Time Grade recommender algorithm on the basis of Grecs agricultural e-commerce system, build and evaluate Grecs recommender system. When Expert Item CF- Time Grade algorithm is designed, This system used a lot of objects such as Hash Map, Hash Set, Tree Set. The time complexity of querying and inserting data is about O(1).Based on the above situations, main purpose of Grecs Agricultural Recommender System is to improve the recommended accuracy of product, speeding up the efficiency of agricultural product recommender algorithm.