Random Forest-based Approach for Classifying Customers in Social CRM

Abstract

Social Customer Relationship Management (social CRM) has become one of the central points for many companies seeking to improve their customer experience. It comprises a set of processes that allows decision-makers to analyze customer data in order to launch an efficient customer-centric and cost-effective marketing strategy. However, targeting all potential customers with one general marketing strategy seems to be inefficient. While targeting each potential customer with a specific strategy can be cost demanding. Thus, it is essential to group customers into specific classes and target each class according to its respective customer needs. In this paper, we develop a Random Forest- based approach to classify potential customers into three main categories namely, prospects, satisfied and unsatisfied customers. The proposed model has been trained, tested, and compared to some state-of-the-art classifiers viz., Artificial Neural Network (ANN), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN) based on several metrics including accuracy, sensitivity, specificity, false positive rate, and false negative rate. The reported results were satisfactory with an accuracy of 98.46%, a sensitivity of 97.69%, and a specificity of 98.84%.