IBM researchers in San Jose have come up with new software that uses the unique ways individuals sign their names, such as their hand strokes and pressure exerted on the writing pad, to determine the validity of signatures and to detect forgery. If a signature is a 95 percent match of the one stored by the software, a green check mark appears, but one that is rated only 1 percent or 2 percent gets a red “x” and is rejected as a forgery. The software, dubbed Sign and Go, will be marketed to retail outlets looking for better ways of detecting identity fraud and will allow retailers to set their own policies on how closely a signature must match the data stored in its network, says IBM researcher Thomas Zimmerman. For example, he says a store may tolerate just 50 percent accuracy for a lower priced item, but require 80 percent accuracy for more expensive items such as washing machines. However, Zimmerman says, “If you require a very high threshold–95 percent–you’ll get a lot of unhappy customers.” Analyst James Van Dyke says it is still unclear to IBM just how big the market will be for this kind of verification software, since a growing number of consumers are shifting to debit card sales based on consumers’ personal identification numbers. Still, Zimmerman says signature verification technology is likely more socially acceptable than such biometric systems as fingerprint verification, and many stores already have electronic touchpads as well as the necessary hardware and software to support the new technology.
Read more here.