Exploring the Impact of Time Spent Reading Product Information on E-Commerce Websites: A Machine Learning Approach to Analyze Consumer Behavior.

This project studies how the amount of time shoppers spend reading product information on e‑commerce sites affects whether they buy the item. Using machine‑learning methods, the system looks at user‑session data, especially dwell time (how long a page stays open), scroll depth (how far down the page the user scrolls), and how they interact with product descriptions. First, clickstream data are cleaned and prepared. Then feature engineering creates temporal features that capture how engaged a user is. These features are fed into classification models—Random Forest, Support Vector Machines, and Gradient Boosting—to predict purchase intent from reading behavior. The research gives step‑by‑step guidance for building predictive models that help e‑commerce sites

improve their interfaces and the layout of product information. Model performance is evaluated with precision, recall, and F1‑score, showing that timing and interaction data are strong indicators of consumer decisions. The implementation also supports real‑time session‑analysis pipelines, providing a practical way to understand digital shopper engagement without using invasive tracking methods.

Message: Hi FE, I need help with “Exploring the Impact of Time Spent Reading Product Information on E-Commerce Websites: A Machine Learning Approach to Analyze Consumer Behavior.” in “Machine Learning & Web Analytics”