Segmentation of Services Provided by E-Commerce Platforms Using PAM Clustering

Abstract

Today, data analytics have helped in solving many critical issues in various domains. From credit card fraud detection, detecting cancer to sentiment analysis, data analytics has come a long way. One of the important domains is e-commerce. With everything going digital, shopping has made its significant place in the digital world. A Partition around Medoids (PAM) clustering algorithm is discussed for grouping the services provided by e-commerce websites to their customers. The PAM clustering method helps to work upon real-time data with mixed data types. The clusters formed will provide better insights and patterns for the businesses to make better decisions in order to improve their customers’ engagement and multiply their profits using the historical data. With limited time and resources to invest, this method helps the company to proceed further by understanding the focus points of each service and their scope of development. Implementation of the proposed PAM clustering is done on the artificial dataset and is found to be effective.