Internet Commerce Security Laboratory (ICSL)

Fraud detection

Fraud detection is a difficult problem, akin to finding a needle in a haystack. It can be difficult to determine the cost benefit of increasing the rate of detection of fraud, which can increase false alerts. We have developed unique methods for improving upon the existing fraud detection methods, using advanced analytics and ripple down rule-based systems that include a "brittleness" test, where if the system is not sure of an answer, it can request specific advice and incorporate that new knowledge.