Bank Customer Value Classification Model Based On The Behavior Pattern Of Groups

The behavior patterns of community groups can reflect the characteristics of time and space.With using the community attribute,the behavior patterns of community group and the property of the bank to make correlation analysis and obtain bank field indicator,it can construct bank customer value division model which can help bank to divide customer better.The work of this paper is as follows:First,a bank customer value division model based on behavior patterns of groups is proposed.The process of model establishment includes feature set selection,which includes community attributes,behavior patterns of community groups and bank domain characteristics,the correlation analysis,as well as the construction of customer value clustering model.Second,I propose the bank customer value division platform design.The platform overall design includes the framework design,flow chart design,and interface design.I give the module design,including community data acquisition module,community group behavior model characteristics,bank data acquisition module,correlation analysis module,and bank customer value division model module.Third,the experimental verification and evaluation of bank customer value division model is proposed.The experimental verification includes the data collection description,the correlation analysis experiment and the bank customer value partition model verification.The correlation analysis experiment includes the correlation analysis between community attributes and behavior patterns of community groups,the correlation analysis between community attributes and bank domain characteristics,the correlation analysis between behavior patterns of community groups and bank domain characteristics.In this paper,the model of bank customer value based on the behavior pattern of community groups has certain significance to the customer value division.At the same time,the platform design of bank customer value partition play a role in promoting for the subsequent development.