Machine Learning Based Product Comparison for E-Commerce Websites

Abstract

Online shopping through e-commerce has gained widespread popularity among consumers, revolutionizing the operations of businesses in the global market. This paper examines the benefits of e-commerce, such as its convenience and the ease of comparing prices and products, as well as the difficulty customers may encounter when selecting the optimal product. To overcome this difficulty, the paper suggests a real-time online consumer behavior prediction system that anticipates a visitor's purchasing intent using session and visitor data and assesses the effectiveness of Continuous Learning with the Naive Bayes strategy. The article also focuses on developing a recommendation system that strikes a balance between increasing precision and safeguarding users' privacy, utilizing the Prize dataset to assess the system's accuracy. Additionally, the paper delves into the domain of opinion mining, outlining its objectives and responsibilities, such as anticipating sentiment, summarizing aspect-based sentiment, and predicting the helpfulness of online feedback and reviews.