MINNEAPOLIS – FICO said its FICO Labs team has been awarded four new patents by the U.S. Patent and Trademark Office, including one for technology that mines transaction data for predictive trends.
The technology, which is behind the Data Spiders module in FICO Model Builder, is used by FICO clients in a variety of applications, such as improving targeting for cross-selling and conversion, improving fraud detection, and predicting online purchase decisions.
The second patent provides an architecture for optimal strategy design using both historical data and human expertise. This architecture defines decision models as a visual influence diagram, translates the influence diagram to a nonlinear optimization problem and derives decision rules that achieve mathematically optimal strategies.
The other two new patents are for new technologies which bolster FICO’s fraud detection and prevention capabilities. FICO received a patent for a fast, accurate “fuzzy” matching technique that can quickly and accurately determine if a given computer-readable record is represented, by exact match or substantially close match, in a large collection of computer-readable records. This is useful in detecting identity fraud. The final patent is for technology used to identify credit card or similar fraud based on account-level factors rather than transaction-level factors.
