3G Subscriber Behavior Model And Empirical Research Based On Data Mining For The Precise Marketing

Since the development of3G services, there was a series of new problems and new phenomena. The increased number of operators and the homogenization of the resources make the3G market enhancing their competition. With the expand number of the customer and the development of individual needs, it increases the cost of the implementation of the marketing strategy. Therefore, the operators have to improve the precision of the management and pay attention to subscriber behavior characteristics, and make rationally use of modern information technology to design more efficient and accurate marketing strategy.For some of the new problems of the3G business development, this paper adopts a combination of theoretical analysis and empirical research, and builts subscriber behavior model based on data mining technology under the guidance of precision marketing. The main content is as follows:1. Under the guidance of precision marketing, the paper improves measure index of the subscriber’s life cycle with3G subscribers, and as a basis to draw the3G subscriber’s life cycle curve, dividing the3G subscriber’s life cycle stage. It refines the marketing aspect and bases theoretical foundition for the accurate observation of the3G user behavior characteristics.2. Based on behavioral economics and the theory of value, the paper builds the feature model of subscriber’s behavior from three aspects of the subscriber’s decision-making behavior, using behavior and loyalty behavior. Creatively put forward the subscriber’s decision-making behavior measure index system, it is described from the value algorithms and mental accounts. Combined with the real data of the subscriber behavior, the paper proposes to measure indexes of use value rate.3. Based on model of subscriber behavior, this paper has two empirical studies between the subscriber identification and churn warning. KGRC algorithm used to solve two questions of k-means algorithm:how to select k value and how to select the initial cluster centers. It helps us identify and improve the3G subscriber label management based on three-dimensional grid of the subscriber segmentation. The model of subscriber behavior can improve explanation on subscriber behavior by the churn rules too.This research provides the necessary quantitative theoretical support in the telecommunications industry applications for the data mining technology. It provides decision-making basis for the telecom marketing strategy options, and enhances the efficiency and effectiveness of marketing strategy.Therefore, it has a great significance in theoretical research and the practical application.