Fraud Indicators Applied to Legal Entities: An Empirical Ranking Approach
Legal persons (i.e., entities such as corporations, companies, partnerships, firms, associations, and foundations) may commit financial crimes or employ fraudulent activities like money laundering, tax fraud, or bankruptcy fraud. Therefore, in the Netherlands legal persons are automatically screened for misuse based on a set of so called risk indicators. These indicators, which are based on the data obtained from, among others, the Netherlands Chamber of Commerce, the Dutch police, and the Dutch tax authority, encompass information about certain suspicious behaviours and past insolvencies or convictions (criminal records). In order to determine whether there is an increased risk of fraud, we have devised a number of scoring functions to give a legal person a score on each risk indicator based on the registered information about the legal person and its representatives. These individual scores are subsequently combined and weighed into a total risk score that indicates whether a legal person is likely to commit fraud based on all risk indicators. This contribution reports on our two ranking approaches: one based on the empirical probabilities of the indicators and the other based on the information entropy rate of the empirical probabilities.
KeywordsProbability Density Function Risk Score Information Entropy Money Laundering Risk Indicator
Unable to display preview. Download preview PDF.
- 1.Verwer, S., van den Braak, S., Choenni, S.: Sharing confidential data for algorithm development by multiple imputation. In: Proc. of SSDBM 2013 (2013)Google Scholar
- 2.Herziening Toezicht Rechtspersonen, Verscherping van het toezicht op rechtspersonen, Ondernemingsstrafrecht en Compliance (February 2012), http://www.boekel.com/media/509294/nieuwsflits_februari_2012_-_verscherping_van_het_toezicht_op_rechtspersonen.pdf.
- 3.Grobosky, P., Duffield, G.: Red flags of fraud. In: Trends and Issues in Crime and Criminal Justice, vol. (200). Australian Institute of Criminology (2001)Google Scholar
- 4.Freelik, N.F.: High-risk profiles and the detection of social security fraud. Journal of Social Intervention: Theory and Practice 19(1) (2010)Google Scholar
- 5.Bonchi, F., Giannotti, F., Mainetto, G., Pedreschi, D.: Using data mining techniques in fiscal fraud detection. In: Mohania, M., Tjoa, A.M. (eds.) DaWaK 1999. LNCS, vol. 1676, pp. 369–376. Springer, Heidelberg (1999)Google Scholar
- 6.Choenni, R.: Design and implementation of a genetic-based algorithm for data mining. In: Proc. VLDB 2000, pp. 33–42 (2000)Google Scholar
- 7.Taylor, C.: Composite indicators: reporting KRIs to senior management. RMA (Risk Management Association) Journal 88(8), 16–20 (2006)Google Scholar
- 8.Peng, H., Gates, C., Chris, S.B., Ninghui, L., Qi, Y., Potharaju, R., Cristina, N.R., Molloy, I.: Using probabilistic generative models for ranking risks of android apps. In: Proc. of Computer and Communications Security, pp. 241–252. ACM (2012)Google Scholar
- 9.Bishop, C.M.: it Pattern Recognition and Machine Learning. Information Science and Statistics, vol. 1, p. 740. Springer, New York (2006)Google Scholar