Several terrestrial services exist for flagging suspected credit-card transactions; one of the largest is that of the Falcon system by the HNC Corporation. The Falcon system incorporates proprietary and neural-network technologies and is the flagship service of HNC, a company that began as a neural-network company in the mid-1980s. Today, Falcon monitors over 240 million payment card accounts, or about 80% of all credit cards in the United States. Falcon uses neural-network-based cardholder profiling and real-time transaction-scoring techniques. HNC started out selling neural network software; however, it quickly found that companies were not interested in buying pattern-recognition programs. Instead, they were looking for business solutions based on the use of this technology. HNC decided to sell fraud-detection services to banks and other issuers of credit cards.
Financial firms found it difficult to administer and maintain their own neural-network models and instead opted to outsource via a subscription service to deal with their credit-card fraud problem. There are several advantages to this business model. For example, new scams can be recognized much more quickly by a fraud-detection expert than by an independent bank or merchant. Fraud detection service providers can also pool both data and technologies via a subscription consortium network, enabling them to detect criminal activity and patterns much more effectively. However, for those interested in learning about fraud-detection techniques, or at the least incorporating some preventive technology on their site, the following case study is provided.