| Customers Relationship Management (CRM) is a kind of advanced thought. It requires the enterprises cooperate with the process outside or inside and improves the service level based on the direction of the customers. CRM is a management software and technology too. It combines the best business practice, data mining, data warehouse, one to one marketing, SFA and other information technologies with together and provides the enterprise with an automatic operation scheme. Based on it, the enterprise has an interface to face it's customer and achieve it's purpose of transforming from traditional enterprise model to modem enterprise model based on E-Business.The data mining is a process of distilling the information and knowledge that is hided in the data sets and isn't aware before, but is useable very much. Perhaps, the data sets include noise, missing and inconsistent parts. The data mining is a new technology that can distill hidden, predictive information from a large database or data warehouse. It can mine the potential model and find the most valuable information to direct the business behavior or assist the science research with something.Based on current data mining research of CRM, combined with the high-power data mining technology of Genetic Algorithm and Neural Network, and based on a lot of English reference, we get some harvest such as:1. The data mining's origin background, the base concept, the technology characteristic and the general process are expatiated first. As the pattern's quality is the precondition of getting credible knowledge model, the data pretreatment should be done before the data mining. In this paper, we propose a scheme to solve the problem of outlier data patterns that the value of outlier data pattern is deleted first and then is predicted by the model of Genetic Algorithm and Neural Network. The result of experiment shows that this method is feasible.2. The BP Neural Network is used widely, but it has several inherent flaws that are disadvantage in the application of CRM. So, researching new optimization methods for BP Neural Network is a important task of this paper. We propose a Genetic Algorithm with a three-level-structure chromosome to optimize the BP Neural Network's topology structure and weighting space at the same time. This is a theory innovative point of this paper. The result of experiment shows that this scheme is feasible. A BP algorithm with smoothing factor is proposed to optimize the weighting space. The result of experiment shows that this scheme is useful for improving the BP Neural Network's generalization ability. This is a technology innovative point of this paper.3.In the CRM, classifying the customers is meritorious. In this paper, we propose a scheme based on Genetic Algorithm and BP Neural Network to classify the customers. This scheme takes full advantage of the BP Neural Network's model identifying ability. The result of experiment shows that this scheme is feasible. Compared with the classical model of decision tree, this model has more excellences such as being easy to build the model and to expand, a great capability of fault tolerance.4. The purpose of CRM is to attract and dig up customers. In the modem E-Business, the individuality marketing is an important research task. We propose a collaborative filteringCRMalgorithm to mine the most possible commodity item that the customer is most favorable. We discuss how to get some important parameters' values of collaborative filtering algorithm through experiment. The experiment shows that this scheme is feasible.5. One of the tasks of CRM system is provide some predictive information for marketing and decision-making. We design a model based on BP Neural Network with a smoothing factor to predict the market share of the enterprise product. The result of experiment shows that the forecasting precision of this model is higher than the classical model. Compared with the classical state-space model, this model is more flexible and has higher forecasting pr... |
PDF Full Text Request