International Journal of Scientific Research and Engineering Development

The volume of online transactions has expanded dramatically due to the e-commerce platforms' rapid expansion, making them an ideal target for fraudulent activity. Because traditional rule-based systems frequently fall short in identifying intricate and dynamic fraud patterns, detecting fraudulent transactions in real time has become a crucial challenge for enterprises. In order to detect e-commerce fraud in real time, this study suggests a machine learning-based method that examines transaction data to spot questionable activity. To differentiate between legal and fraudulent transactions, the system trains classification models using data preprocessing techniques. A web-based application framework is created to effectively monitor transactions and forecast fraud. The suggested approach assesses transaction characteristics and looks for irregularities that might point to fraud. According to experimental data, the system can detect possible fraud more accurately and quickly than traditional approaches. This strategy can help e-commerce companies increase consumer trust, decrease financial losses, and improve transaction security. The study emphasises how machine learning techniques can be used to create scalable and sophisticated fraud detection systems for contemporary digital commerce environments.

K Naresh, Kuntrapakam Dhananjayulu,"Machine Learning-Based Real-Time E-Commerce Fraud Detection" International Journal of Scientific Research and Engineering Development, V9(2): Page(1446-1450) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.