FICO Patents 100th Analytic Innovation

New patents cap 20 years of breakthroughs that have changed the way businesses make decisions

MINNEAPOLIS--(BUSINESS WIRE)-- FICO (NYSE: FICO), the leading provider of analytics and decision management technology, today announced that it has been awarded 8 new patents by the U.S. Patent and Trademark Office, bringing its portfolio total to 100. The breakthroughs covered by these patents are behind FICO's industry-leading solutions in credit scoring, credit account management, credit fraud management, insurance fraud management and other areas.

Since 1990, FICO's commitment to research has increased the momentum of its analytics innovations. About one-quarter of the 100 patents were awarded to FICO in 2010 alone.

FICO's patents span predictive analytics, fraud management, insurance fraud management, business rules and credit technology. They have resulted in the development of a number of products recognized as the leaders in their respective categories, including the following:

• The FICO® Score - launched in 1989 and continually updated based on FICO's advances in predictive analytics - is the world's most used credit score, with more than 10 billion sold each year in 20 countries. The FICO Score is the standard measure of U.S. consumer credit risk, used by lenders, regulators, rating agencies and consumers.
• FICO™ Falcon® Fraud Manager - based on neural network models and cardholder profiles, protects more than 2 billion bank cards worldwide, and has saved issuers more than $10 billion in the U.S. alone.

The new patents cover breakthroughs in predictive analytics, fraud detection and business rules management. These patents are for technologies and methodologies that:

• Compress consumer profiles used in fraud detection, to improve processing while improving fraud detection accuracy
• Process combinatorial optimization problems, and in particular, co-clustering optimization problems, for applications such as natural language processing and bio-informatics, using so-called "simmered greedy optimization"
• Provide comprehensive protection from identity fraud with a detection system that uses neural networks to identify potentially fraudulent cases
• Reconcile multiple interrelated rule sets in a workflow model of a business process
• Transform large or complex decision trees into compact, optimized representations to ease viewing and interaction by a user
• Graph trends in massive databases
• Detect insurance premium fraud or abuse using predictive software
• Use consistency modeling methodologies to identify potentially fraudulent or abusive activity in healthcare, both by providers and patients.