The Design And Implementation Of Intelligent Order Analysis And Logistics Delay Prediction Platform

Amazon market has been expanding in recent years,and the logistics services provided by its supply chain have a huge demand in various countries.It outsources the distribution services through its own logistics distribution base station^[1].According to the distribution process of logistics,the data analysis team of supply chain needs to decompose,analyze and improve them one by one.With the increase of logistics content,how to provide a more stable and efficient logistics analysis integration tool for the team is an urgent problem.The Amazon intelligent order analysis and logistics prediction platform designed in this paper mainly provides the data analysis team of supply chain with the functions of visual analysis and persistent storage of real-time order data,integration,browsing and subscription of various data reports,and delay prediction and display of logistics time.The system is based on Amazon redshift big data warehouse,and uses SSM framework?i.e.the integrated architecture of spring MVC,spring and mybatis?as the development mode to develop and implement the back-end.The front and back end of the system interact with each other through JSON,which is realized by four functional modules:user information management module,real-time order analysis module,data report management module and logistics delay prediction module.Because there are many data sources in the system,lambda,dynamodb,redshift and My SQL are respectively used to store the system data,cache the real-time data stream,persist and generate the chart.In addition,the author analyzes,compares and tests the weather data preparation phase of the delay prediction module in detail,so as to achieve a better result The author and his team choose the decision tree model as the training model of the system by comparing the decision tree algorithm with the multiple regression model algorithm and testing the accuracy of the results.In the development process of the project,the author mainly participated in the development of the back end of the system as well as the later testing and maintenance work.At present,the system has completed all development work and testing,and all functional modules can be used normally and stably,which can help data analysts to view data reports more intensively and predict the possible logistics delay more accurately.In the future,the system will continue to expand its functions and optimize its performance.