RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

IRIS Fraud Prevention

Product
Developers: Iris analytics
Date of the premiere of the system: 2013/10/29
Branches: Financial services, investments and audit
Technology: Cybersecurity - Fraud detection system (fraud)

IRIS Fraud Prevention is technology of fraud prevention.

The antifraud system considers each transaction sent to a processing center and classifies it using a rule set, the formulated system developers on the basis of knowledge of methods of fraud and statistics on such transactions. Rules can be simple ("the client cannot spend money somewhere, except the country of accommodation") and difficult ("if the client withdrew money in the ATM in the country of accommodation, it cannot earlier, than in five hours, withdraw money in other country"). The it is more difficult than the rule, the it is less than false operations when the bank refuses to the client accomplishment of legitimate transaction, and the more rules need to be set. In the analysis of transaction about several tens parameters can be checked — the system is more perfect, the she is capable to consider bigger number of signs.

Initially in IRIS no rules are set. At implementation the history of transactions of a system for the period is loaded into it, among these transactions there has to be a certain number of the confirmed fraudulent courses. A system analyzes operations on 50 different parameters and makes own rule set. These rules are formulated in human language, i.e. any operator will be able to learn for what reason a system did not miss a certain transaction. Than longer the specific copy of IRIS works, the more transactions is exposed to the analysis, the higher it efficiency.

"By our calculations, IRIS correctly defines 85% of frodovy transaction, and only 0.2% of transactions in a false manner are defined as frodovy. Actually on each detected fraud we have one, or it is even slightly less, false operation. At the same time for the traditional systems of 12 false operations on one detected fraud it is considered normal. Some number of false operations is anyway inevitable as clients can and make not intentionally the actions similar to a fraud, but them not being" — Igor Korolev, the promotions manager of software solutions of Wincor Nixdorf Russia company told.

According to the statements of developers, IRIS it suits for use practically in any financial industry. Moreover, it can define cases not only a fraud, but also, for example, unfair fund raising for charitable needs or even drug traffic (if the seller uses for it any certain e-wallet or the bank account).