A framework for data mining‐based anti‐money laundering research
Date | 15 May 2007 |
DOI | https://doi.org/10.1108/13685200710746875 |
Pages | 170-179 |
Published date | 15 May 2007 |
Author | Zengan Gao,Mao Ye |
Subject Matter | Accounting & finance |
A framework for data
mining-based anti-money
laundering research
Zengan Gao and Mao Ye
School of Economics and Management,
Southwest Jiaotong University, Chengdu, People’s Republic of China
Abstract
Purpose – The purpose of this paper is to propose a framework for data mining (DM)-based
anti-money laundering (AML) research.
Design/methodology/approach – First, suspicion data are prepared by using DM techniques.
Also, DM methods are compared with traditional investigation techniques. Next, rare transactional
patterns are further categorized as unusual/abnormal/anomalous and suspicious patterns whose
recognition also includes fraud/outlier detection. Then, in summarizing the reporting of money
laundering (ML) crimes, an analysis is made on ML network generation, which involves link analysis,
community generation, and network destabilization. Future research directions are derived from a
review of literature.
Findings – The key of the framework lies in ML network analysis involving link analysis,
community generation, and network destabilization.
Originality/value – The paper offers insights into DM in the context of AML.
Keywords Data analysis,Money laundering, Crimes
Paper type Research paper
Introduction
Money laundering (ML) is the processing of criminal, “dirty” money to disguise their
illicit origin and make them appear legitimate and “clean.” IMF Managing Director
Michel Camdessus estimated in 1998 that between 2 and 5 percent of the global GDP is
laundered annually. Evidence also shows tha t ML finances terrorist attacks
worldwide. So anti-money laundering (AML) research is of critical significance to
national financial stability and international security.
ML behavioral patterns and ML network structural features are essential to AML,
but traditional research focuses on legislative conside rations and compliance
requirements. It is methodologically limited to incident identification, avoidance
detection, and suspicion surveillance. Investigations are generally manual, tedious,
time-consuming, and resource-intensive. So challenges are often made to their high
false positive rate (FPR) and inefficiency with voluminous data sets.
Comparatively, data mining (DM) can reduce data preparing time, determine
detection priority, improve FPR, and lessen the pressure of manpower, training, and
budget. It proves particularly effective and efficient in:
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1368-5201.htm
The research is accomplished during the author’s visit to UQ Business School, the University of
Queensland, Australia. The author’s grateful thanks are due to Professor Peter Green,
Dr Dongming Xu, and Dr Jon Heales for their invitation and support.
JMLC
10,2
170
Journal of Money Laundering Control
Vol. 10 No. 2, 2007
pp. 170-179
qEmerald Group Publishing Limited
1368-5201
DOI 10.1108/13685200710746875
To continue reading
Request your trial