Research On Vehicle And Cargo Matching Management In Logistics Information Platform

With the rapid development of the logistics industry,the logistics information platform has been established in this context.In recent years,the problem of vehicle and goods matching management has made great breakthroughs in the theoretical research and practical application.However,the logistics information platform still faces the cold start problem of new drivers and the sparseness of cargo scores.In view of the above two problems,the main research work is summarized as follows:1)The cold start problem of new drivers: A cross-integration of driver attribute characteristics and cargo attribute characteristics is proposed,and a new driver-car match matching management model based on the random walk PersonalRank algorithm is used to train the popular list.The method utilizes the attribute characteristics of the driver and the cargo,extracts behavior data according to the attribute characteristics of the cross combination,and trains to obtain a popular list.Experiments show that the method plays a positive role in improving the cold start problem of new drivers.2)The problem of sparseness of cargo scores: The feature engineering consisting of driver attribute characteristics,driver historical behavior characteristics,driver real-time behavior characteristics,cargo attribute characteristics,cargo historical behavior characteristics and real-time behavior characteristics of cargo is proposed.Then the XGBoost sorting algorithm is used to train the optimal personalized vehicle matching management model,and the experimental results show that the method can significantly improve the matching accuracy.3)Based on the A Logistics Information Platform,the vehicle and cargo matching management system including the data model,the new driver matching model and the old driver matching model was designed and implemented,and finally the retention rate of the new driver of the platform and the matching accuracy of the old driver were improved.