Abstract:

Modeling the customer behavior becomes a key issue in customer retention, direct marketing and product promotion. Many data mining techniques have been designed to address the problem of capturing the true behavior of customers for various businesses applications, in which the data can be represented in form of demographic and transactional datasets. In this paper, we propose to use association rules mining technique for modeling customer behavior in form of constructing generalized profile association rules.Our proposed scheme based on transactional databases that consist of demographic data (i.e.,personal information), and transactional data where items bought in a customer transaction can come from any level of hierarchy/ taxonomy representation (i.e., computer is-a Hardware). The constructed profiles can be used for segmenting the customers into groups, and identifying the potential ones for targeted marketing and Customer retention. We implement pruning techniques and interestingness measures to select the relevant and interesting rules for profiling process. Experimental results on a synthetic dataset are provided to show the effectiveness of the proposed scheme for generalizedcustomer profiling.

Keywords:Generalized customer profiles,association rules mining, customer modeling

(Full Paper 466KB)

BibTex:

}

(Full Paper 466KB)