Proactive Approaches to Detecting Fraud

1
Proactive Approaches to Detecting Fraud
2
Data Mining Software

• An inductive approach
• Looks for anomalies in databases.
• Limited ability to handle large databases and
  refined queries.
• May be used to
• Examine trends in purchases (vendors, amounts,
  timing)
• Examine trends in sales and returns (by customer,
  by sales clerk, timing)

3
Digital Analysis

• Benfords Law
• An inductive approach which looks for anomalies
  in the first digit of numbers.
• Works with naturally occurring data.
• May be used to identify populations where fraud
  may be occurring.
• Accounts receivable balances
• Sales amounts

4
Benfords Law
5
Deductive Fraud Detection

• Begins by identifying which type(s) of fraud may
  be occurring.
• Must understand the business
• Must anticipate types of frauds
• Must determine symptoms
• Uses databases and information systems to look
  for symptoms of the suspected frauds.
• Follows up on symptoms.

6
Analysis of Financial Statements

• Vertical analysis
• Horizontal analysis
• Ratio analysis
• Note that all of the above require an
  understanding of normal relationships in a
  business.