Fachartikel
Enhancing Target Group Selection Using Belief Functions

Enhancing Target Group Selection Using Belief Functions

One of the most critical decisions in customer relationship management, interactive marketing, and in direct marketing is the assignment of customers to groups with similar response characteristics. This decision usually is based on previous purchase behavior and socio-demographic variables. Unfortunately, information from different sources is likely to provide the managers with conflicting indications of group memberships. In this study we utilize Multi Sensor Clustering and a Rough Sets approach for assigning individuals to groups. We outline the combination of different assignment recommendations by means of the Evidence Theory and the Transferable Belief Model. For an empirical application we rely on a set transaction data recorded from a customer loyalty program.