Self-Payment Fraud Detection on Automated Teller Machine

Volume 1 - Issue 1

Mohammad Hossein Barkhordari*

Received: September 22, 2018;   Published: September 28, 2018

DOI: 10.32474/CTCSA.2018.01.000102

Full Text PDF

To view the Full Article   Peer-reviewed Article PDF

Abstract

Over the past decade the amount of transactions and reported frauds on Automated Teller Machines (ATM) has significantly increased. Various types of frauds have been reported around misusing ATM cards and many methods have been deployed to detect and prevent them. In some countries, banks sell ATMs to investors under predefined circumstances and pay them in commission in order to increase the availability of the service but some ATM owners have been found to create fake transactions to obtain extra commissions. This paper attempts to detect such frauds using a two-stage method. In the first stage fraudulent customers are detected by certain rules and in the second stage their accomplices are identified using transaction loop and cycle detection algorithm. Transactions of an Iranian bank have been used to evaluate the proposed method and all detected fraudsters by system were confirmed by bank fraud detection office.

Keywords: Fraud Detection; Cycle Detection; ATM Fraud Detection; Data warehouse

Abstract| Introduction| Related Works and Definitions| Proposed Method| Results| Conclusion| References|