An unsupervised purchase-based customer clustering method for e-supply chain

Making use of clustering technology to segment customers properly is the most important problem. The two most frequently used algorithms are the K-mean and the SOM algorithm. In this paper, a novel unsupervised clustering technology-ISODATA algorithm is proposed for the customer segment based on the customer´s purchasing behavior. Unlike the K-mean algorithm, the clusters are merged if either the number of members in a cluster is less than a certain threshold or if the centers of two clusters are closer than a certain threshold in our method. On the contrast, the clusters are split into two different clusters if the cluster standard deviation exceeds a predefined value and the number of members is twice the threshold for the minimum number of members. It has some further refinements by splitting and merging of clusters. The customer clustering will be illustrated through a case study on the e-commerce database of bookshop.