The development of e-commerce significantly contributes to technological advancement, especially for businesses adopting the concept. The growth of e-commerce has seen a significant increase, reaching 196.47 million users in 2023. In e-commerce, a wide range of product variations is provided to users, which can lead to errors or confusion in product selection. Product categorization is crucial in e-commerce to assist users in navigating efficiently. However, manual categorization is less effective as it can be time-consuming. This study aims to clarify the factors of concern in grouping using the K-Nearest Neighbors (KNN) algorithm in product categorization on the e-commerce platform. This research focuses on whether the novelty lies in the implemented algorithm, the variables used, or the applied grouping parameters. This work applies the XGBoost algorithm to improve the effectiveness of product categorization in e-commerce through ensemble learning approaches. The research findings indicate that boosting algorithms like XGBoost outperform individual algorithms like KNN regarding classification accuracy. This proves that ensemble learning approaches may greatly enhance product classification in e-commerce. The testing process of the implemented e-commerce system in this study also provides confidence in the theoretical and practical benefits of applying this research to enhance efficiency and user experience in product categorization on the e-commerce platform.