Research Of Data Mining For Supporting Cross-marketing Of Financial Products

Today, financial enterprises at home and abroad are trying to transform their concept of marketing from "product orientation" to "customer orientation" in order to achieve greater competitive advantage. Cross-marketing could provide a set of solutions of products or services to customers for the purpose to reduce the marketing cost, expand sale, increase profit, and enhance customers' loyalty. Wherefore, the research of cross-marketing becomes one of hotspots for many scholars at home and abroad in recent years. However, because financial data have the characteristics such as large quantity, uncertainty, etc, the current cross-marketing activities are carried out under the lack of scientific analysis of customers and products. The cross-marketing schemes are proposed based on many assumptions, thus are of less relevance, low efficiency, and have less practical value.The author believes that to carry out efficient cross-marketing should be on the basis of the two important aspects which are customer segmentation and products association analysis. Customer segmentation can help enterprises analyze the various types of customers' consumption characteristics, and form marketing strategies for different customer groups and establish the basis for cross-marketing. Products association analysis can help enterprises understand the potential relationship between the various types of products, and make strategies to implement cross-marketing according to customers' historical records of transactions.First, the theoretical basis of data mining technology is introduced, the status of research and application of the data mining technology in the financial enterprises is described, and the existing problems are analyzed. The K-means clustering algorithm is chosen to establish the customer segmentation model, and the optimized association rule Apriori algorithm is chosen to establish the product association model after comparing many algorithms. Dan Pelleg and Andrew Moore's idea of improving the K-means clustering algorithm is adopted in order to increasing the stability and the rationality of the clustering results for customer segmentation and to improve the efficiency of the algorithm. Then, the customer segmentation model and the products association analysis model are combined together to support cross-marketing, and a technology solution supporting cross-marketing for the financial products is put forward. Finally, a decision-making supporting system based on the proposed solution is designed and a simulative application process is described.