Abstract :

Online shopping behavior analysis has become increasingly essential due to the rapid growth of e-commerce platforms and evolving consumer preferences. Predicting a shopper’s intention level, such as purchase, browsing, or abandonment, helps organizations enhance personalization and improve marketing strategies. Machine learning techniques enable accurate modeling of user behavior using features like session duration, page views, cart activity, demographics, and transactional history. This work proposes a predictive framework that classifies user intention levels using supervised learning models. The proposed system integrates preprocessing, feature extraction, and classification to improve decision-making accuracy. Experimental results indicate significant performance improvements compared to traditional analytical methods. The model demonstrates strong reliability and supports real-time decision recommendations for e-commerce systems.