Font Size: a A A

Rearch And Implementation Of Intelligent Recommender System Of Agricultural Resource Supply And Demand Information Based On User Role

Posted on:2017-01-09Degree:MasterType:Thesis
Country:ChinaCandidate:H Y WeiFull Text:PDF
GTID:2348330488980035Subject:Agriculture
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet technology,the quantity of network information has grown explosively.It is considerable difficult for farmers to seek what they need and hard for supplies to sparkle what their intend in the agricultural trading platform.Currently,providing information service to meet the personalized requirements,which are aimed at users demand characteristics constituted by different roles,regions and reasons,has become the urgent problem in agricultural e-commerce siteAs an important personalized service model,Collaborative filtering recommendation system has become more and more widely used in the field of Internet.This paper developed a recommendation pesticides system,and then combined the intelligent recommendation technology with agricultural trading platform to design and implement agricultural supply-demand information intelligent recommendation system based on users' roles.Compared with the traditional algorithm of recommendation,recommendation algorithm based on user roles integrated agricultural seasonal,regional and features,which is more suitable for agricultural resources recommended.The main contents of this paper are as follows:(1)Research on the current development situation of the existing recommendation system and the recommendation algorithm.This section elaborates the relevant fundamental theory of content-based recommendation algorithm,collaborative filtering algorithm and so on,and summarizes the application in agricultural trading platform for personalized recommendation algorithm by combined with agricultural recommendation status.(2)Research on collaborative filtering algorithm based on improved.Firstly,this section establishes I-U score matrix correction via calculating the I-I similarity matrix and collect user behavior.And then it constructs user similarity matrix U-U and calculate the user similarity utilizing Pearson correlation coefficient so that we can determine the nearest neighbor.Finally,it products the last generation prediction score and recommendations.(3)Realization of agricultural supply and demand information intelligent recommendation system.This section on the basis of the establishment of inter-entity E-R diagram,design the system data table detailedly.By building storm distributed real-time computing framework,it design and empolder agricultural supply-demand information intelligent recommendation system,which is consist of agricultural purchase,personalized recommendation,order management,and basic query.Under the support of Anhui Province 12 th Five-Year science and technology research projects — “Research on the key technology of agricultural logistics informationization and application oriented course of electronic commerce”,this paper applies collaborative filtering recommendation algorithm based on user roles to the agricultural products personalized recommendation service successfully,reduces the user's search time and promotes the completion of the transaction.
Keywords/Search Tags:collaborative filtering, similarity calculation, user roles, recommendation
PDF Full Text Request
Related items