Application Research Of Customer Segmentation Based On Clustering Under Railway Freight Big Data Platform

In recent years, the railway freight information construction in China has made great breakthroughs, but a lot of freight data is lack of effective management and utilization, so it has great value to carry out the technology of big data in railway freight business data mining research. Customer segmentation is the basis of freight transport marketing, and it has the advantage of identifying customer groups and configuring enterprise resources, which can lead to more profits for the enterprise. But the current railway freight customer segmentation method is based on experience and simple statistical division, it can neither accurately distinguish the type of customers, nor effectively support the marketing decision. This paper makes improvements of RFM model, which is commonly used in customer segmentation. At the same time,this paper combines the RFM model with clustering algorithm to provide new solution in railway freight transportation with massive complex data.In this paper, the main work has the following aspects.(1) According to the character of railway transportation, this paper makes improvement on the traditional RFM model, and proposes the KFM model.(2) The traditional K-means clustering algorithm is sensitive to the initial clustering center and is easy to fall into the local optimum, due to low accuracy of customer classification. In order to improve the segmentation accuracy, this paper puts forward the improved K-means clustering algorithm. And the experiments show that the improvement is successful.(3) This paper combines the KFM model with the K-means algorithm, and it carries out the customer segmentation with the data in Railway E-commerce System. The result has fully showed the characters of different customers, which makes up the shortcomings in traditional RFM models.(4) In the Hadoop big data platform, this paper realized the parallelization for data standardization and K-means clustering based on MapReduce, which improved the performance of the algorithm. It is capable of a large number of data processing tasks.In this paper, the clustering mining technology is used in customer segmentation under railway freight big data platform. This method determines the type of customers with different values and behaviors, shows the customer categories for enterprise. This can improve the accuracy of marketing decision of freight department.