Research On Customer Group Consumption Behavior Based On Clustering And Association Rules

Research on the commonness and individuality of customer consumption behavior, is an important prerequisite for making marketing decisions and strategies. With the advent of the era of big data, every hour and moment, consumer does produce a large amount of data, data mining of the behavior of customer indicates that the analysis of behavior must be potentially valuable, however, in order to explore more information using massive amounts of data, large data mining research on consumer behavior has existing problems, this paper is based on the background of telecommunication industry and based on the existence of the telecom industry, found that the homogenization of competition, the growth of operators benefit is slow; the number of the stock of customers is becoming smaller and smaller, and the quality of service for the key customers is in falling; the phenomenon is obvious, resulting in waste of resources, low efficiency. In order to achieve the optimal allocation of social resources, to seek the most reasonable consumption of resources, get the most social benefits, and provide personalized consumer demand for mining, as much as possible and to stimulate consumer demand potential, providing a variety of services and products, to meet the diversified market demand, enhance the efficiency of enterprises at the same time, provides the high quality service for the key customers accurate.In this paper, from the data point of view, first of all the research background and significance, and literature review, determine the research topic, based on the research topic, research methods for clustering large data processing on the one hand, on the other hand, the method of association rule mining user cross behavior, two aspects to explore the related theory and method. Then build customer consumption behavior of combination model, a telecom operator as an example, the application of consumer behavior model of consumer behavior analysis, finally, put forward some marketing strategy suggestions, finally is a summary and outlook. The results of this study are as follows:(1) The relevant theoretical results, explores the evolution direction of data mining technology, analyzes the advantages and disadvantages of clustering algorithm and application in data clustering, the principle and classification of association rules algorithm, summarize customer group segmentation.(2) The method of high dimension reduction is proposed for the correlation coefficient method based on the vector cosine angle and Pearson, to construct a partition clustering method in data processing model, and a large amount of data in high dimensional data clustering has good effect, can accurate customer segmentation, and according to the clustering results, explore the characteristics of customer consumption behavior.(3) For the study of customer's characteristics of crossing behavior, to build crossing t consumer behavior model of association rules mining, mining methods and ideas put forward customer cross behavior characteristics of single dimensional and multidimensional association rules combined.(4) The advantages of combining the models of the two methods of clustering and association rule, build customer consumption behavior based on combination model, and the research on customer consumption behavior of this model is applied to a mobile operator, The model is used to study common behavior and individual behavior of customers in two aspects of consumer behavior.(5) customer consumption behavior through the case study of a Mobile Corporation, and constructs the index system of the consumer behavior of a Mobile Corporation, to provide basis for the behavior indicators of customer segmentation, customer consumption behavior from different perspectives, different customer market, analysis of customer consumption behavior characteristics with groups, mining crossing consumption behavior characteristics of key customers group, and finally puts forward targeted marketing strategies and suggestions based on clustering and association rules.From the analysis, theoretical results in the telecommunications industry, analysis of customer consumption behavior has important practical significance and practical value, research ideas and methods for a large data mining, as well as user centric era, the importance of model combination method of customer behavior, at the same time, the massive data of specific market background processing and customer consumption behavior characteristics, has the very good reference significance to study cross behavior characteristics of key customers.