Discovery of Identity Fraud Through Data Mining and Visualization, 10-R9688

Printer Friendly Version

Principal Investigators
Danny T. Lents
Michael Magee

Inclusive Dates:  02/20/07 06/20/07

Background - Identity theft has been the fastest growing white-collar crime for the past seven years. Approximately 10 million Americans fall victim to ID theft annually. There is significant state and federal interest in combating identity theft.

Identity thieves often work as part of a network of thieves operating over multiple jurisdictional boundaries. They have the advantage of making it difficult for law enforcement to identify patterns of activity that may expose them.

Approach - The goal of this project was to investigate data mining and visualization of data collected from ID theft victims, businesses, and police reports.

This study was divided into the tasks of data identification, data preparation, data visualizations and experimentation. The project team worked with several government agencies to learn the types and uses of ID theft related data they collected. Most agencies were willing to identify the types of data collected, but none of the agencies was able to provide sanitized sample data. This required creating data for software development and testing.

The project team concentrated on four general collections of data based on the analysis of data collected by the agencies: data identifying the victim, data identifying the suspects, data collected by businesses about the victim, and data collected by businesses about the suspects.

The project team looked for ways to implement data queries and visualizations that might uncover clues for an investigator in common ID theft scenarios. Identity theft cases reported in the news were also examined to provide ideas for scenarios to analyze.

Accomplishments - A prototype user interface called Crime Scope was developed that provides a tool for an investigator to analyze ID theft data. The prototype provides a means to demonstrate ideas to potential clients interested in analyzing ID theft data.

2007 Program Home