Application Of Data Mining Technology In Automobile Insurance Telephone Sales

Auto insurance telephone sales is a new marketing model,which is based on the telephone as the main means of communication,with the help of network,fax,SMS and mail.With the popularity of this marketing model,a large number of auto insurance customers and product sales business data are preserved.This large number of business data often exists in different systems,the compatibility between them is poor and there is no unified planning.Besides these,there are still a lot of interference data.With the deepening of the information industry and the fierce competition between enterprises,how to use the existing data to study the marketing model which is suitable for the auto insurance telephone sales has become a difficult problem for the insurance companies.The technology of data mining is to extract implicit,unknown,potentially valuable information from large scale data,which can solve this problem very well.The focus of this paper is setting up a data warehouse suitable for auto insurance combined with the needs of the insurance company,based on the characteristics of cross-system,poor compatibility and many interfering data,through the understanding of the auto insurance business,designing a customer segmentation model based on cluster analysis and a model based on customer classification on this basis,achieving the classification of customers and product sales forecast by using SAS data mining tools and developing a different marketing strategy for each type of customers from the purchase probability and the purchase price so as to guide the salesman to sell targeted.The establishment of data warehouse,the understanding and selection of data mining algorithms and the cleaning of auto insurance data are the main tasks of this paper in research and development.This thesis,in-depth understanding of the field of insurance industry knowledge,has a detailed description of the application of data mining technology in the auto insurance telephone sales.The main work of this thesis is listed as follows: 1,Starting from the auto insurance business,this thesis introduces the design and implementation of the conceptual model and logical model of the customer data warehouse of auto insurance;2,This thesis introduces the algorithms of data mining,such as K-means algorithm and logistic regression algorithm;3,This thesis introduces the process of the empirical analysis of insurance data using data mining technology.From the results of research and development,some conclusions are obtained which are helpful to the auto insurance decision in this thesis.For example,the low renewal rate and high yield customers are often high-end cars and car transfer,this kind of customer is easy to lose,we can put the offer to retain customers,finally premium renewal rate increased by about 2%-3% through these marketing strategies.Besides,targeted recommendation of auto insurance during automobile insurance telephone sales can increase the premium renewal rate.This exploration provides some reference value for the application of data mining technology in the auto insurance telephone sales.However,we need to design the attribute values of related topics according to the data acquisition in practice.