The Research And Implementation Of Customer Segmentation System Based On Clustering

Customer segmentation refers to a process that divide customer sets into one or many classes using customer attributes data. Most customer segmentation methods do not make full use of customer data. This paper mainly research on the Customer segmentation method combing customer static attributes and dynamic attributes data based on clustering, as well as the design and implementation of the prototype system based on the method.Cluster analysis is one of the main functions of Data Mining and Knowledge Discovery, which groups data area sets into classes by nature and gives a description of features for each class. This paper introduces a Domain based Modified Clustering Algorithm (DMCA), based on semantic distance, which can automatically determine the number of clusters, and can work well on the datasets that are described by mixed categorical and numerical attributes. This paper presents a MACSM (Mixture Attributes Customer Segmentation Model) based on the DMCA algorithm, integrated with both customer static and dynamic attribute data,a more comprehensive analysis of customer characteristics, improve the customer segmentation accuracy. For customer dynamic attribute data, the eigenvalue extraction method is proposed, a method for the discretization of continuous attributes data based on domain background knowledge is also introduced, the model gives out the interpretations of the clustering with the conceptual complex expression. Finally, the model has also proposed a similar RFM model of customer value analysis. Experiments show DMCA algorithm is effective and MACSM model can improve customer segmentation accuracy.