Retailer Segmentation via NeXClass Ordered Multi-criteria Classification Methodology and System

Abstract

NexClass is a multi-criteria methodology and decision support system developed to support multi-criteria classification into predefined nominal (non-ordered) categories. In this paper we introduce an extension of NexClass methodology at classification problems ordinal (ordered) categories. For each category an entrance threshold is defined by decision maker, which indicates its boundary with respect to the evaluation criteria. Then, assignment of alternatives into the classes is based on the concept of non-exclusivity. This defines at what degree an alternative can be included into a specific category. Alternatives are evaluated according to the evaluation criteria, and non-excluding degrees are calculated for each category. Finally, alternatives are assigned into the category for which non-excluding degree gets the minimum value. The NexClass decision support system implements the above classification algorithm, providing a user-friendly interface. This work presents the methodology and the classification model along with a real world application to retailer segmentation. Findings show that the methodology is valid and can be used for segmentation problems in marketing or relevant domains.