On The Understanding of Customer Purchase Behavior: A Case Study on the Computer Systems E-Commerce Website

Abstract

The tremendous growth of the E-commerce industry over the last decade drew attention to the importance of understanding customers' behaviors and modeling user-system interactions to maximize customers' satisfaction by providing them the best user experience and gaining new customers to the e-commerce website. Within the scope of this study, software architecture _that can help us understand customers' purchase behaviors and then make a binary prediction based on their interactions with the system to determine whether a computer configuration will be bought_ is proposed and explained thor-oughly. For this purpose, machine learning algorithms are utilized to correctly model complex user-system interactions. A prototype of the proposed software architecture is developed using the latest technologies, and experimental studies are performed on top of this prototype in order to understand how successful our software architecture is. The conducted experimental results prove the efficiency of the proposed software architecture in real-world data, with 83% and 88% scores for the accuracy and F1-score metrics, respectively.