Cracking Multi-Account Fraud: Data-Driven Solutions for E-Commerce Platforms

Abstract

Multi-account fraud poses a significant challenge for e-commerce platforms, enabling users to exploit promotions, evade platform defenses, and re-enter systems after being restricted or suspended. This paper presents a comprehensive study on data-driven methods to detect and mitigate multi-account fraud. By leveraging advanced analytics and machine learning techniques, we aim
to identify all accounts linked to the same user, providing a holistic evaluation of user behavior and risk. Our approach focuses on preemptive strategies to thwart fraudulent activities, thereby preserving marketing budgets and optimizing promotional efforts.
We propose a robust framework for e-commerce companies to implement effective multi-account policies, enhancing their risk management capabilities and safeguarding the integrity of their platforms.