Using Class Membership based Approach to Improve Predictive Classification in Customer Relationship Management Systems

Abstract

Recently, the diversity of data collected on both social networks and digital interfaces is extremely increased. This diversity of data raises the problem of heterogeneous variables that are not favourable to classification algorithms. Although machine learning and predictive analysis have significantly improved the efficiency of the classification in customer relationship management (CRM) systems, their performance remains very limited by heterogeneous data processing. In this paper, we propose a new predictive classification approach well adapted for targeting actual CRM systems. Our approach consists of preprocessing each type of feature and constructing a reduced array. From this reduced array, the class membership computations become very faster and perform the predictive targeting of a new instance great accurately. The results of the experiments carried out on four types of data from the CRMs showed that the proposed algorithm is a good tool for strengthening these systems not only to optimize their loyalty actions but also to efficiently acquire new customers.