Research On Customer Relationship Management Methods For Railway Container Transport

Railway container transportation has the advantages of large volume,fast speed,and convenient connection of multiple transportation modes,playing an important role in sea rail intermodal transportation,international railway intermodal transportation,and modern logistics.The railway department continuously strengthens its freight marketing methods,adapts to the market-oriented operation of railway freight,and attaches increasing importance to customer relationship management.At present,research and application of customer relationship management are mostly concentrated in industries such as telecommunications and finance,while there is less research on customer relationship management in the field of railway freight transportation,especially in the field of railway container transportation.This article focuses on customer relationship management in railway container transportation,proposing models and methods for three key issues: customer value measurement,customer segmentation,and customer churn prediction and warning.It deepens existing research results and provides theoretical basis for railway container transportation customer marketing.The main research works as follows.(1)Analyze the main characteristics of China’s railway container transportation market and customer relationship management.Elaborate on the current situation and development trend of China’s railway container transportation market from aspects such as domestic railway container transportation and international railway intermodal transportation.Analyze the OD distribution and flow characteristics of railway container transportation in China.Elaborate on the shortcomings of customer relationship management in China’s railway container transportation from two aspects: methodology and information construction,and propose the development requirements for customer relationship management in railway container transportation.(2)Analyze the main influencing factors of customer value in railway container transportation from three aspects: customer natural attributes,current value,and potential value.A customer value evaluation index system for railway container transportation is established,and a customer value calculation method is proposed.From the perspectives of comprehensive value and itemized value,combine with k-means clustering algorithm,customer segmentation models for railway container transportation are established,dividing customers into 4 and 8 categories,and proposing principles for developing differentiated service strategies.Considering that random selection of initial clustering centers for k-means may lead to slow convergence speed and insufficient accuracy of clustering due to insufficient number of clusters,local density values and attraction range indicators in Density Peaks clustering are introduced to assist in selecting initial clustering centers,thereby improving the efficiency and accuracy of the k-means clustering algorithm.(3)Construct a system of customer churn prediction and warning methods,propose monitoring indicators for customer churn prediction and qualitative indicator discrimination methods,form a customer churn monitoring indicator system,introduce data mining methods such as CART decision tree and BP neural network to construct a customer churn prediction model,set prediction model evaluation indicators to compare the accuracy of the model,and finally propose a combined model of Lagrange function based on multiple prediction models to increase the accuracy of customer churn prediction.(4)Select customer data from typical railway container transportation enterprises and construct the case study of customer relationship management technology system.Calculate customer value and subdivide customers according to customer value.The two segmentation methods are basically in line with the "Pareto principle" in the big customer theory.Two types of data mining based churn prediction models are used for churn prediction,with prediction accuracy of over 90%.However,due to the small number of sub models,the improvement effect of the combined model is not ideal.Finally,analyze the reasons for customer churn and propose corresponding maintenance strategies.