Design And Implementation Of Urban Traceability Platform For Important Product

Traceability data of a product includes various important data collected in the process of market circulation of the product,including product transaction information,merchant information,inspection information and etc..One of the most concerned topics of the construction of traceability systems is how to conduct management,query,and analysis on the traceability data.This thesis reviews the status quo of the domestic and abroad traceability system construction and the analysis of the traceability data.And constructs an urban traceability platform for important products based on SSM(Spring + Spring MVC + My Batis)framework in order to address insufficiency in the current analysis and management of the traceability data.Then a highlight of this thesis is the proposal of a traceability data analysis model integrating data mining methodology.Innovative work of this thesis includes:(1)Proposal of performing multi-angle analysis on the traceability data using the data mining methodology.Cluster analysis is performed on features of the products using k-means algorithm to generate reasonable clustering results,thereby ensuring reliability and validity of the data.Association analysis is performed on heavy metal pollution detection results using Apriori and FP-growth algorithms to make targeted improvement of the products,thereby ensuring quality safety of the products.(2)Proposal of regression prediction of the traceability data using a method combining PCA(Principal Component Analysis)dimension reduction and SVR(Support Vector Regression).A prediction model is built based on some physical and chemical features of product in the traceability data using SVR.Sensitivity analysis is performed on the features to evaluate importance thereof.Based on the analysis result,dimension reduction is performed on the original data using PCA to identify important features.The final model exhibits good performance with reduced calculation amount.