US 11,875,350 B2
Systems and methods for improved fraud detection
Durga S. Kala, Foster City, CA (US); Kenny Tsai, Foster City, CA (US); Juharasha Shaik, Foster City, CA (US); and Aditi Khare, Foster City, CA (US)
Assigned to Visa International Service Association, San Francisco, CA (US)
Filed by Visa International Service Association, San Francisco, CA (US)
Filed on Sep. 12, 2019, as Appl. No. 16/569,117.
Prior Publication US 2021/0081948 A1, Mar. 18, 2021
Int. Cl. G06Q 20/40 (2012.01); G06Q 20/42 (2012.01); G06N 20/20 (2019.01)
CPC G06Q 20/4016 (2013.01) [G06N 20/20 (2019.01); G06Q 20/425 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
receiving, at a risk management server from a transaction data server, a plurality of test transactions each including one or more test transaction parameters, wherein a plurality of known fraudulent transactions make up a subset of the plurality of test transactions;
applying machine learning techniques to the plurality of test transactions to identify one or more fraud transaction parameters each having a range of fraud parameter values;
associating, by the risk management server, each range of fraud parameter values with a risk level;
transmitting, by the risk management server, a graphical user interface (GUI) to an issuer server;
displaying, by the risk management server via the GUI:
one or more of the one or more fraud transaction parameters,
one or more selectable risk levels associated with each respective fraud transaction parameter, and
one or more selectable threshold fraud rates indicating a proportion of past merchant transactions known to be at least one of fraudulent, with high frequency merchants, or with high risk merchants;
receiving, by the risk management server via the GUI, a user selection of one or more of the one or more risk levels associated with each respective fraud transaction parameter;
receiving, by the risk management server via the GUI, a user selection of a threshold fraud rate of the one or more selectable threshold fraud rates;
applying, by the risk management server, the user selection of the one or more risk levels and the threshold fraud rate to the plurality of test transactions to determine a fraud detection rate; and
displaying, by the risk management server, the fraud detection rate on the GUI, wherein the method is performed using one or more processors.