An Integrative Framework for Recognizing Fraudulent Behavior in Shared Online Exchanges

The main goal of transaction security systems has always been to detect and stop fraudulent transactions on e- commerce platforms. However, it is difficult to catch offenders using only previous order data because of the secret nature of e-commerce. Many research have attempted to create technologies that prevent fraud, but they have failed to take into account the dynamic behavior of users from various angles, which results in ineffective fraud detection. This study offers a revolutionary fraud detection method that combines process mining and machine learning models to track user activity in real time in order to address this problem. First, we create a process model for the B2C e-commerce platform that includes user behavior detection. Second, an anomaly analysis technique that can identify noteworthy aspects in event logs is introduced. The collected features are then fed into a classification model that uses Support Vector Machines (SVM) to identify fraudulent activity. We show through experiments how well our approach captures dynamic fraudulent actions in e-commerce platforms.