An analysis of structural changes of customer segments by a hybrid method of clustering and association rule

Abstract

Customer segmentation is one of the main and key issues in customer relationship management studies. One of the main challenges of customer segmentation is segments’ instability and changes over time. Segments’ changes can be categorized into two main categories including structural and content changes. The focus of this paper is on the structural change which is very important. To analyze the structural changes, a hybrid method of clustering and association rule mining techniques is presented and implemented on the data of a main service provider in telecommunication industry in Iran. Based on the results, the structural changes are categorized into cyclic changes and changes with trend which are discussed. The results show that a growing trend has been formed implying creation of a new segment in this corporation. The results provide the marketing managers with useful knowledge to make more effective marketing decisions.

Keywords

• association rule mining
• Clustering
• customer segmentation
• data mining
• Structural changes