 |  | | | Benefits 
Find underlying similarity despite variations that fool other software
Make more accurate decisions about potential fraud, with less manual intervention |
| |  |
Overcome Inconsistencies that Fool Other Software
The software that does the best job of finding records is also the best software to detect many of the most common types of fraud. People who perpetrate fraud have learned that it doesn't take much to fool most computer systems. A rebate is limited to one per customer? No problem: "Richard Ballinger" and "Dick Bollinger" could fool your system, but still be close enough that the bank happily will cash the multiple checks you issue to this unscrupulous customer. Most fraud is detectable if your software perceives similarity the way humans do... Perpetrators often still need to receive mail, deposit checks, remember the information they provided you, or preserve "plausible deniability" ("Oh, gee, it must have been a typo"). So fraudulent claims or registrations will usually contain a large core of similar information. People who perpetrate fraud learn from experience how to introduce just enough inconsistencies to fool your system. Netrics makes this game much more difficult: the Matching Platform automatically perceives similarity much the way humans do. Out of the box, it matches information that is similar, even if it isn't exact. It comes up with more candidate transactions which might be fraud. You can decide how much you want to automate the process, and can even "train" the system to discriminate fraud based on examples of how your own experts do it. There is no complex rules-base to build or maintain. The bottom line: find more fraudulent transactions with fewer mistakes and less need for manual intervention.
|
