SOURCE: S3

S3

September 10, 2010 06:00 ET

S3's Advanced Matching Software Approaches Calculations of the Human Mind

TeraMatch® 2.0 Data Quality Software Focused on Eliminating False Positive Matches

AUSTIN, TX--(Marketwire - September 10, 2010) -  After 8 years of deployment in Fortune 50 enterprises and constant refinement by developers, S3's TeraMatch® 2.0 has moved a significant step closer to eliminating false positive matches when comparing large databases. TeraMatch® combines a series of proprietary and publicly available algorithms to calculate a "similarity score," which approximates the choices of human intelligence.

"Obviously, you can't beat the human mind," said S3 CEO Jack Holt. "But by refining our algorithms with years of data analysis our software is now able to discern the finest distinctions in matches. And, of course, the human mind can't match millions of records in seconds."

A recent study conducted by S3's MIT scientists compared three lists of 559 names with additional fields like addresses and social security numbers. There were 79 duplicates entered into the databases and each of those had a slight modification designed to be difficult to detect. TeraMatch® 2.0 found all 79 duplicates and made only 3 false positive matches.

"We ran the TeraMatch® 2.0 test against the government's software that's being used on some of the terrorist databases," said S3 CTO Mark Davies. "Their algorithms only found 17 of the 79 duplicates and it also turned up 97 false positives. Databases will always have mistakes so the matching software has to keep improving, which is what ours is doing."

TeraMatch® 2.0 is presently deployed to prevent fraudulent providers from enrolling in Medicaid and has been used to identify and match telecom circuits and asset databases to create more accurate inventories and billing.

About S3 - (www.S3.com) is an Austin, Texas based company focused on providing business crisis solutions for network outsourcers, financial services, and healthcare industries.

Contact Information