Research Of Customer’s Feature Extraction Based On The Analysis Of The Clients’consumotion Behavior

With the constant development of economy and globalization, the future market competition will be more tend to focus on service competition. As the increasingly difference and personalized customer demand and the formation of buyer’s market, the enterprise must follow the idea "take the customer as the center" when they provide customer services and products. If the enterprise wants to base in the future market competition, the key elements are to use new technology, new ideas for service innovation to improve the service quality and to satisfy customers’ increasingly changeable and persounalized needs. In the customer behavior analysis field, how to innovative application of new scientific and cultural knowledge and some advanced technology idea to analyze customer behavior has become a practical problem in the world today. However, facing the increasingly the difference and personalized of customer demand and the increasing mass of data of customer, it has been in a white-hot stage both from the technical level or in service now.In this case, how to deal more effectively with customer data, analyze customer behavior and resolve customer characteristics become one of the key challenges for customer behavior analysis.This paper discusses the association rules in customer relationship management application. Based on customers’ consumption behavior of customer groups subdivision, can make the enterprise according to the customer value levels of different decide how to in the client distribution enterprise limited resources, and then according to the different needs of customers, the design and implementation of the different customers maintain strategy. Its purpose is to keep firmly to the enterprise that part of the most value for the customer, and the potential the current low value customers in the future into high value customer, encourage those who either now or in the future of the company is no value for the customer to its competitors, and in the end reach enterprise of the overall profit maximization. Association rule method can help enterprise more scientific and effective achieve this goal. In this paper, the market marketing theory, put forward the auto after-sales service enterprise customer consumption RFP model, through the AHP method got the auto after-sales services RFP index proportion, get used to cluster analysis is the basic data of weighted index RFP, and application of improved ant colony clustering method combined to customer RFP classified data. Then use improvement association rules algorithm analysis of the characteristics of all kinds of customers ontology, realize customer ontology according to the purpose of feature extraction. In the model and the algorithm was put into practice after, this paper for a car service company of data into the empirical study, this paper the whole content of the practical application of, did it from the customer behavior index to feature extraction process from customers of a seamless joint. At the same time, this article to one automobile company service information data into the life of empirical studies, confirmed the suggested method and the model is practical and effective, and have a strong usability and enforceable, thus up to the unity of theory and practice.