UA Professor Presents Developing Fraud Detection Technology to U.S. Congressional Committee
This month, Joe Valacich of the Eller College of Management presented his early stage company's new technology for identifying suspicious behaviors indicative of fraud and deception to a Congressional Committee on Homeland Security hosted by U.S. Congresswoman Martha McSally.

By Raymond Sanchez, NASA Space Grant Intern, University Relations - Communications
July 29, 2015


Whether through an e-mail phishing scheme or a hacked social media account, most Internet users have likely had at least one experience with digital deception.

This is a particular problem in the business world, where online employee fraud costs businesses across the globe $3.5 trillion annually. For governments, detecting digital deception is not only a matter of efficiency, but of national security.

On July 24, Joe Valacich, a University of Arizona Eller College of Management professor, presented his early stage company's new technology for identifying suspicious behaviors indicative of fraud and deception to a Congressional Committee on Homeland Security. The presentation was hosted by U.S. Congresswoman Martha McSally in Washington, D.C. 

Neuro-ID's software can identify suspicious behaviors based on a computer or smartphone user's typing, scrolling or mouse movements across a screen. The technology could be invaluable to government and a wide range of industries including insurance, pharmacy, healthcare and e-commerce.

"The possible applications for this technology are limitless," McSally said. "Any process that uses an electronic form, whether it’s for background checks or visa applications, carries the potential for deception, and this technology would help our government personnel better detect that deception and make smarter decisions."

Neuro-ID was born out of a desire to merge the science of human-computer interaction and big data technology.

From online shopping to surfing the web, it should come as no surprise that our Internet activity leaves a distinct digital fingerprint, which reveals a lot about our preferences and behaviors. Usually, Internet companies will track the browsing habits of users, and use that data to improve their service or offer personalized recommendations. Valacich argues that by doing this, companies are getting a limited picture of online behavior.

"'Personalization 1.0' is looking backwards at a user's history to learn more about them, but 'Personalization 2.0' is all about determining what they're doing right now," said Valacich, a professor in Eller's Department of Management Information Systems. "We wanted to understand a person’s intentions at a given moment. With such insight, you can provide a better service encounter in an electronic commerce context, or determine that someone is not being honest when completing an online application form."

In 2012, Valacich and his then-graduate student Jeff Jenkins, now a professor at Brigham Young University, attended a conference where scholars stressed the need for the next generation of polygraph testing and threat detection.

During a break in the meeting, Valacich and Jenkins discussed a neuroscience study that showed a person's cognitive processes have a strong influence on the nature of their hand movements. Inspired by the research, the pair recognized the potential to use the computer mouse in an inexpensive, scalable solution for identifying suspicious behaviors indicative of deception.

"When people are thinking about doing something deceptive, they have cognitive conflict that's reflected in their physiology and behavior," Valacich said. "Are they hesitating more? Are they frequently changing mouse or scroll directions? Are they getting slower and less accurate with their clicks? These all indicate unusual behavior, and they are all things we can detect."

Since 2012, Valacich and his team have developed dozens of features for identifying suspicious behaviors, including mouse acceleration, area of movement and deviations from an idealized mouse trajectory. All of the features are weighted and combined into a Suspicion Index Score™, which measures the likelihood that a user is engaging in fraudulent behavior. The process has been tested in dozens of scientific studies involving thousands of human subjects.

With the help of Tech Launch Arizona, Neuro-ID has secured Valacich and Jenkins' U.S. patents developed from their research. Valacich emphasized that TLA has been critical to his company's successful launch.

"David Allen and the entire Tech Launch team have been wonderful to work with," said Valacich. "They're invaluable partners in commercializing this technology. It's a very professional and effective way of achieving a common goal."

At the moment, Neuro-ID is offering web-based screening that utilizes keyboard, touchpad and mouse movements to gather information. Software that analyzes touchscreen usage is currently in the works, which will bring the technology to smartphones and tablets within the next month or so.

"I’m proud to be able to showcase this technology, its potential uses within our government, and the great innovation happening at the University of Arizona," McSally said.

Extra info

Read more about Tech Launch Arizona's successes at UANews.org by reading: UA Marks Record Second Year in Research Commercialization, Several Notable Start-ups

Share