State of the Art Literature on Anti-money Laundering Using Machine Learning and Deep Learning Techniques

Abstract

Money laundering (ML) is the process by which criminals convert illicit funds gained from drug sales, human trafficking, and other illegal operations into legal funds that can later be spent freely and without restrictions. The annual amounts of money involved in ML are in the billions of dollars, and it is believed that it accounts for 2% to 5% of worldwide GDP [10]. Governments around the world are trying to combat this type of criminal activity by requiring banks and other financial institutions to report any such activity. In this paper, we will conduct an in-depth examination of the current literature on anti-money laundering, focusing on the areas of machine learning, deep learning, data mining, and big data. After reviewing the papers, we noticed that while various graph algorithms are being used, such as graph embedding and centrality algorithms, there is a shortage of community detection algorithms that could help identifying organizations and groups involved in money laundering. This gap in the literature suggests a potential area for future research to investigate the use of community detection algorithms in anti-money laundering systems to enhance their ability to detect illegal activity.