Business Process Mining for Internal Fraud Risk Reduction: Results of a Case Study

Everybody can recall some kind of fraud that has been all over the news. If it were Enron, WorldCom, Lernout & Hauspie, Ahold, Société Générale or another case does not matter. Fact is that fraud has become a serious part of our life and hence a serious cost to our economy. Several studies on this phenomenon report shocking numbers: forty-three percent of companies worldwide have fallen victim to economic crime in the years 2006 and 2007 [1]. The average financial damage to companies subjected to this survey was US$ 2.42 million per company over two years. Participants of a study of the Association of Certified Fraud Examiners (ACFE) estimate a loss of five percent of a company’s annual revenues to fraud [2]. Applied to the 2006 United States Gross Domestic Product of US$ 13,246.6 billion, this would translate to approximately US$ 662 billion in fraud losses for the United States only. These numbers all address corporate fraud. There are several types of corporate fraud. The most prominent distinction one can make in fraud classification is internal versus external fraud, a classification based on the relationship the perpetrator has to the victim company. Management fraud is an example of internal fraud, where insurance fraud is a classic example of external fraud. In this paper we present and apply a framework for internal fraud risk reduction, where risk reduction stands for both fraud detection and prevention. In a previous paper, we already presented a framework with data mining being the core of that framework. In this paper we complement that framework with a process mining part (see Figure 1). Process mining aims at uncovering a process model based on real transaction logs. This relative new research domain can be applied in several ways for the purpose of internal fraud risk reduction.