E-commerce Personalization Revolutionized by Birch and FCM Clustering

Abstract

Understanding customer behaviour in the fast-paced e-commerce environment of today is essential for online success. E-commerce platforms collect enormous amounts of clickstream data, which may be used to reveal consumer preferences and improve the online shopping experience. In order to segment and categories clients based on their clickstream data, this study offers a novel approach (BFC-HT) that combines Birch clustering, Fuzzy C-Means (FCM) clustering, and dynamic hyperparameter tuning via Grid Search and Random Search. In a highly competitive online world, these analytics enable e-commerce businesses to personalise marketing initiatives, improve user interactions, and eventually promote increased consumer happiness and conversions. In doing so, our research significantly advances the rapidly developing fields of e-commerce analytics and customer segmentation by providing a solid framework for extracting useful information from the vast amount of available clickstream data.