Actualidad -- Julio 14, 2006

CSC'S FRAUD VISION HELPS BANKS FIGHT CRIME

Check fraud is a costly plague for consumers and banking institutions. In its most recent survey, the American Bankers Association found $5.5 billion in annual attempted fraud, resulting in actual dollar losses of $677 million. Despite continued vigilance by banks, increasingly sophisticated fraudulent checks still slip through the cracks. But armed with CSC's FraudVision, a suite of automated, patent-pending software tools, banks can fight fraud better than ever before, reduce their losses and increase customer confidence.

FraudVision was designed specifically for today's high-volume, image-based check processing operations. When Check 21 was signed into law in 2004, banks were no longer required to ship paper checks among themselves. Instead, the image of the check is now considered legal tender and can be transmitted electronically. While the benefits to banks include quicker, more efficient check processing, many traditional fraud detection methods were rendered ineffective.

"Traditional analyst methods involved things like feeling the paper, sniffing the ink and bending corners, a lot of very tactile, upfront check examination techniques that you can't do with images," says Chaz Engan, Fraud Solutions Development manager, who led the FraudVision development team. He notes that FraudVision counteracts the loss of the actual check with enhanced detection tools.

Taking the holistic view
Building upon traditional fraud-detection tools, such as signature verification and data analysis, FraudVision takes a holistic view of the check, rather than focusing on one single factor. It analyzes such characteristics as check stock, signature, payee name and handwriting, extending intelligent character recognition to include both machine and cursive handwriting. FraudVision maintains a profile of these characteristics, so when an image of a check is presented, the system automatically determines if the check characteristics match. Using a Combined Analysis Rules Engine (CARE), FraudVision combines the results of this image analysis with results from external data analysis systems, resulting in better quality and more accurate results.

For example, if the bank's data analysis system flags a check because of a suspect dollar amount, FraudVision can quickly verify the check characteristics to rule out fraud. Likewise, if the data analysis system misses a fraudulent check because the payment pattern matches the user's profile exactly, FraudVision can detect that it is indeed fraudulent because the check stock is irregular. This image and data analysis is combined and presented to the user at a single, easy-to-use Fraud Analyst Workstation.

"It really is a new approach to solving a business problem," says Jerry Blodgett, CSC's director of product management for Integrated Payments. "Traditionally, there have been signature verification companies, and companies focused on the data of the check payments, for example, the check number. But there was nobody taking the comprehensive view of looking at all the different pieces of the check. That is the breakthrough thinking of FraudVision."

Technical excellence
FraudVision was built upon more than 15 years' experience in pattern recognition, and is part of CSC’s suite of applications that automates transaction processing for banking operations. For their work on FraudVision, the team of Engan, Katerina Blinova, Bob Nugmanov and John Lee received the 2006 Award for Technical Excellence, CSC's top honor for innovation. Also contributing were Pierre Luu and Joanna Willis.

"This team's achievement demonstrates CSC's ability to develop innovative software to solve an industry-wide problem," says Jim Cook, president of CSC's Financial Services Group. "FraudVision delivers value to financial institutions, commercial businesses and consumers."

More analysis, better results
With a large volume of checks to analyze daily, it is important to banks that only the most suspect checks are referred to investigators for analysis. "A bank that processes a million checks a day doesn't have time to have someone look at every check. They want to exclude 99 percent of incoming checks as assumed good," Engan says.

FraudVision's extremely low false positive rate helps banks save on labor costs, because fewer checks are sent to an investigator, and those that are sent are more likely to be counterfeit or forged. FraudVision is so effective, Engan says, that it allows for human analysis of every suspect item, even in high-volume operations.


Related Information:

Learn about the 2006 CSC Award for Technical Excellence.

Read more about FraudVision.

See more of CSC's banking solutions.

Contact us and let our experience help you produce results.