Research Of Potential Customer Finding Of Railway Freight Based On Data Mining

With the rapid development of high-speed railway in China, railway transportation achieved the separation of passenger transport and railway freight gradually, its transport capacity has been greatly increased. However, in the fierce competition of freight market, market share of rail freight declining day by day and the gap of market share between rail freight and road freight becomes bigger. Finding an effective way to identify the potential customers is very urgent under the situation of increasing traffic capacity and declining market share, to avoid wasting of traffic capacity and then enhancing their market share through developing new customers. It can efficiently identify the railway freight customers’characteristics by analyzing and digging out its clients’information and trading record based on data mining technology, refining some valuable information in the widely spaced data, so that provide the decision making basis for railway freight companies with these useful information.Firstly, this paper analyzes the current situation on potential customers identifying of railway freight, and comes up with an overall scheme on potential customers identifying based on product characteristics and customer demand characteristics. Secondly, we discussed the related data collect and pre-process scheme of potential customers identifying in detail, constructing a data warehouse on rail freight potential customer identification. Furthermore, the paper constructs rail freight customers identifying models to get customer type segment with existing customers in two dimensions:customer value and demand-product match. Based on the customer segmentation types, this article uses decision tree to extract attribute characteristics of each type of customer and form the customer feature library. Then, a list of potential customers and the corresponding types of these potential customers can be obtained by using the rules of customer characteristics to match the external company data which has been collected. Finally, this article realizes the model of potential customers finding of railway freight using the database tools and data mining tools, selecting the optimum algorithm by comparing the algorithm. Then, a list of potential customers and the corresponding types of these potential customers can be obtained. According to the type of potential customers, railway freight companies can be informed of the possible value obtained from potential customers and the difficulty of customer acquisition. Also, this article proposes a development program of each type of potential customer base on their own characteristics accordingly.In this article, the model of potential customers finding of rail freight can not only differentiate whether the analysis target is potential customers, but also to roughly estimate how much value a potential customer can bring to and it’s difficulty level on acquisition, to provide decision making basis of new customers development.