Data Mining for Fraud Detection

Data Mining for Fraud Detection

Roberto Marmo (Universita' Pavia, Italy)
DOI: 10.4018/978-1-7998-3473-1.ch079
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

With the increased use online and electronic resources both by the companies and the customers the problem of fraud has been rising in the last decade. Data mining to classify, cluster, and segment the data and automatically find associations and rules in the data, that may signify interesting patterns, including those related to fraud. This chapter aims to introduces to the concepts of fraud, processes and tools involved in data mining techniques, as well as the importance, challenges, and use cases.
Chapter Preview
Top

Background

The Association of Certified Fraud Examiners ACFE defines fraud as: “The use of one's occupation for personal enrichment through the deliberate misuse or misapplication of the employing organization's resources or assets”. Therefore, a fraud is a criminal deception, use of false representations to obtain an unjust advantage, or to injure the rights and interests of another.

The main reason behind the commitment of fraud is to achieve gain on false ground by an illegal means. This has a dramatic impact on the economy, law and the human moral values.

There are different kinds of fraud: internal/occupational or external. Internal frauds happen when an employee commits fraud against his or her organization. External frauds involve a wide range of schemes, including vendors, customers or thefts by other third parties.

Frauds can be classified according to three areas:

  • Motive: The main reason behind commit fraud;

  • Means: The nature or form of the fraud involved to achieve the goal;

  • Methods: The facilities and tools that are used to commit fraud.

Diversity of fraud regards organizations, governments, and individuals such as external parties, internal employees, customers, service providers and suppliers.

Each fraud area has its specific characteristics and faces different challenges, therefore it is important to analyze the fraud scenario in order to establish:

  • What is a fraudulent behavior and modus operandi over time;

  • What is a fraudulent person;

  • Degree of available knowledge about known fraud;

  • Quantity and type of available data.

Planning audit strategies is a posteriori fraud detection problem with prevention purpose of analyzing historical audit data and constructing models of planning effectively future audits. A case study is presented by Bonchi (1999) which illustrates how techniques based on classification can be used to support the task of planning audit strategies.

Implementing the right technology is the key factor in order to analyze and to prevent fraud. Fraud prevention and detection are the proper protection mechanism against fraud, but fraud prevention alone is not sufficient, fraud detection is required to protect vital services in business systems. Some of the major benefits of investing in fraud detection software include:

Key Terms in this Chapter

Fraudster: A person who commits fraud, especially in business dealings.

Data Mining: Finding insights which are statistically reliable, unknown previously, and actionable from data.

ACFE: World's largest anti-fraud organization and premier provider of anti-fraud training education and certification, web site http://www.acfe.com .

Fraud: A fraud is a criminal deception, use of false representations to obtain an unjust advantage, or to injure the rights and interests of another.

Financial Statement Audit: The examination of an entity's financial statements and accompanying disclosures by an independent auditor.

Fraud Detection: An automatic system based on description of user behavior and fraud scenario, in order to detect a fraudulent activity as soon as possible to reduce loss.

Complete Chapter List

Search this Book:
Reset