This paper aims at proposing a hybrid mining approach for e-retailers to analyze returns patterns from both the customer and product perspectives, classify customers and products by returns ratios, and then adopt proper returns policies and marketing strategies to the identified classes for reducing returns transactions and sustaining profits. A multidimensional data framework and a process model for the hybrid mining approach are presented with a simulated example for validation.
Relation:
Proceedings of the 9th Joint Conference on Information Sciences, JCIS 2006, Volume 2006, 論文編號 CIEF-113
