Presentation of a combined data mining model using associative rules and clustering to identify the dominant patterns of customer behavior (Case study: Ansar Bank)

Subject Areas : Futurology

iman gharib ¹ , Abbas Toloie ² , Kambiz Heidarzadeh ³

1 - PhD Candidate Industrial Management Department, Science and Research branch of Islamic Azad University, Tehran Iran [e-mail]
2 - Professor industrial Management Department, Science and Research Branch of Islamic Azad University, Tehran, Iran
3 - Associate Professor Business Management Department, Science and Research branch of Islamic Azad University, Tehran Iran

Received: 2018-08-27 Accepted : 2019-01-21 Published : 2019-11-22

Keywords: Customer Dynamic Behavior, segmentation, Association Rules,

Abstract :

Background: A new matter that arises in term of the dynamic behavior of customers is considering the customer segmentation. Based on the banks competitions to increase market share as well as the psychological and environmental factors the dynamics of customers’ behavior should be considered over time. Transferring customers to different sectors over time and discovering the dominant models in their displacements between sectors are of important topics in this context. Objective: this article aims to identify the behavioral clusters, the dominant patterns of displacement, and the leading characteristics and patterns of customer displacements with a focus on the customer dynamics behavior of Ansar banks. Design/Methodology: A Hybrid method based on clustering and association rules has been proposed. Finding: four different behavioral group of customer are identified:" low-value customers with sustainable model", "low-value customer with unsustainable profitability model", "turned away customers with average profitability", "loyal customers with low profitability". Relations between of these groups are analyzed by association rules