Design And Implementation Of Precursor Chemical Supplier Recommendation System

The rapid development of internet technology increases the amount of information people can obtain,and reduces the cost of information acquisition.However,the issue of information overload has also been highlighted.Because of the particularity of the precursor chemical industry,users engaged in it can obtain mass information easily,but it is difficult for them to choose merchants as their suppliers.This thesis focuses on the personalized recommendation technology as well as its practical application,and expounds the development of Precursor Chemical Supplier Recommendation System for relevant users.Relying on the Big Data Intelligent Analysis Platform of Precursor Chemical produced by SBS Innovation Technology Co.Ltd.,in this thesis,a collaborative filtering algorithm based on RFC model and commodity attribute,as well as a recommendation algorithm based on user characteristics,are proposed and implemented.Eventually,the design and development of the Precursor Chemical Supplier Recommendation System are completed.The work and innovation of this thesis are as follows:(1)For users who have purchase records in the data source,a recommendation algorithm based on RFC model and commodity attribute is proposed.In order to adapt to the current data,which does not contain the monetary value,and the present situation of users' purchasing behavior originates from their actual need for a certain chemical,this algorithm adopts a kind of method that obtains users' comprehensive evaluation to the commodity through the improved RFM model,solves the problem that users' implicit feedback is difficult to evaluate.It obtains users' preference to the supplier through their appraisal to the commodity,solves the contradiction which the supplier evaluation is ambiguous.The Item-CF is adopted to reduce the impact of mass users on the recommendation efficiency.(2)For users who have no purchase and sale data in the data source,a recommendation algorithm based on user characteristics is proposed.The algorithm is based on the relationship between users' characteristics and their purchase behavior,obtains their latent preferences,realizes the personalized recommendation for users who have not feedback behavior,and solves the user cold-start problem.(3)On the basis of requirement analysis and researches of recommendation technology,the work of establishing the Precursor Chemical Supplier Recommendation System,which includes the design of general framework,function module and database,has been completed.Based on Oracle database,Java programming language,Bootstrap and jQuery frameworks,ECharts visualization library,it's finally realized the Precursor Chemical Supplier Recommendation System.