Research On Data Mining Of Data For Railway Freight Transportation

With the increasingly intensive competition in the transportation market and the promoting of the informatization process, a lot of original data and information are also accumulated. Railway freight transportation data has the trait such as large quantity, complicated structure, and multi-tiered network. Data mining technologies are needed to be used to research the target customers, life cycle, customer value and development issues. This paper actively explored the way of China railway freight transportation data mining, and research have been done in the following several aspects.The composition, characteristics and hierarchy of railway freight transportation data were analyzed, and railway transportation freight data were collated and classified.By excavating the associative rules between railway freight data, we came to the conclusion that the strategy of centralized processing and optimized loading can be effectively implemented on the railway. By using cluster analysis on the freight data, the paper has concluded that the railway freight has its own target market, thus railway can make its target customer clear. Time series analysis can be applied to predict volume of rail freight scientifically and accurately, which can provide a reference for the rational arrangements of rail transportation and production, the maximized use of rail transport capacity, the formulation of transportation plans, and the compilation of transport scheme. By using case analysis, the paper showed that the ARIMA model can be effectively used for the prediction of rail freight data, and the results have a high goodness of fit.The life cycle of railway customers are researched in this paper, which include division of their life-cycle into different stages, the change of characteristic statistics in different stages, and the evolution of customer loyalty. Based on the above research, this paper proposed the thinking of how to define the different stages of the life cycle of railway freight customers and its relevant model. Based on the characteristics of profit curve in the different stage of customers’ life cycle, this paper established the fitting function of customer profit in the different stage of the railway freight transportation and calculated customer value by using this function. Finally, the subdivision algorithm of the customer value is designed.By taking marketing costs, customer types, expected revenue and other factors into consideration, this paper has constructed development models of railway customers according to its different types, including potential customers model and competitive customer development model and customer retention model. By using case analysis, the paper showed the usability of these models.