Data Mining Research And Application Of Orders In The Tobacco Industry Crm

As the coda in protection period after the accession to the WTO of our country closes on, domestic more and more monopoly industries begin to open the markets progressively. The tobacco trade of China will face opening the markets in an all-round way within a few years in the future too; enter the overall free competition situation of internationalization of domestic market. In order to regard it as the focal point of work of the trade reform to improve customer satisfaction in State Tobacco Monopoly Bureau to the present severe situation, carry out "good service "in the whole course of reforming the work. Adopt advanced customer's relation to manage, classify the customer. It is the important means to improve customer satisfaction to adopt the difference service to the customer of different classifications.Tobacco enterprises have already accumulated a large number of order data and customer's data during the process of dealing in, there are abundant information in the data. Date mining is a technology that it can find hidden knowledge in a large amount of data. We can subdivide the customer by analyzing the order data using clustering method of data mining technology, and then we can provide different service to different customers.In this paper, we subdivide customer group by data mining technology, utilizing the statistics software SPSS, taking the order data and customer information data for data source, using the method of the cluster. Before the data are excavated, carry on the pretreatment of the data to initial data at first. After transforming the initial data, dealing with the wrong value and lacking value, adding information, we get the right data that fit to the data mining. Set up RFM analyzing models, then adopt k-means clustering algorithm. After that, carry on clustering analysis for the three elements-" Regency "," Frequency " ," Monetary ". Then we can get the result of subdividing the customers.After subdividing, the customers are subdivided into four groups. The four groups are called: VIP customer, important customer, ordinary customer, small customer. We can find that these four kinds of customers have obvious differences after analyzing the data, such as geographical position, behavior of ordering the cigarette, the brand that customers order, etc. These differences are the most obvious between VIP customer and small customer. In addition, after analyzing the "Frequency" element of...