Web Mining Techniques - A Framework to Enhance Customer Retention

Abstract

In e-commerce, retaining customers on the web is a difficult task that requires a good understanding of customers' behavior to be able to predict their needs and interests. Web usage mining (WUM), which is the application of data mining techniques to improve business, helps in understanding customers' behavior on the web. Therefore, this paper proposes and implements a framework to enhance the quality of customer recommendations. Providing customers with what they are looking for helps increase their satisfaction, which will lead to improved retention with the company. The proposed framework was tested and evaluated. The result of testing the proposed framework illustrates that the recommendations based on merged techniques (like clustering, classification, association, and sequential discovery) achieve strong accuracy with a precision value of 74%, coverage of 100%, and an average overall efficiency of F-measure of 86%. which means that the merged technique outperformed each technique and attained much higher overall coverage.