Systems and methods for improved fraud detection

1 . A computer-implemented method comprising:

applying, by a risk management server, fraudulent transaction data to a machine learning model to determine fraud transaction parameters associated with known fraudulent transactions;

transmitting, by the risk management server, a graphical user interface (GUI) data signal for accessing a GUI to an issuer server, wherein the GUI is to display selectable:

risk levels corresponding to the fraud transaction parameters;

threshold fraud rates; and

acceptable risk thresholds;

receiving, by the risk management server from the issuer server, a rule selection data signal based on selections at the issuer server via the GUI, the rule selection data comprising:

a risk level selection;

a threshold fraud rate selection; and

an acceptable risk threshold selection;

receiving, by the risk management server, a transaction data signal comprising live transaction data associated with live transactions being processed by the issuer server;

identifying, by the risk management server, first risky transactions from the live transactions based on the risk level selection and the fraud transaction parameters;

identifying, by the risk management server, high risk merchants associated with the live transactions based on the threshold fraud rate selection;

identifying, by the risk management server, second risky transactions from the live transactions, wherein the second risky transactions are associated with the high risk merchants;

analyzing, by the risk management server, the first risky transactions and the second risky transactions based on the acceptable risk threshold selection to identify live transactions that are suspected fraudulent transactions;

transmitting, by the risk management server, a transaction denial signal to the issuer server to cause the suspected fraudulent transactions to be automatically denied;

determining, by the risk management server, a fraud detection rate based on indicative of a proportion or a percentage of the known fraudulent transactions identified as suspected fraudulent transactions;

transmitting, by the risk management server, a fraud detection rate data signal to the issuer server for display via the GUI; and

receiving, by the risk management server from the issuer server, an updated rule selection data signal based on updated selections at the issuer server via the GUI, the updated rule selection data comprising at least one of an updated risk level selection, an updated threshold fraud rate selection, or an updated acceptable risk threshold selection, and the updated rule selection data configured to increase the fraud detection rate.

2 . The method of claim 1 , further comprising:

identifying, by the risk management server, high frequency merchants associated with a transaction total that exceeds a transaction frequency threshold; and

identifying, by the risk management server, third risky transactions from the live transactions, wherein the third risky transactions are associated with the high frequency merchants.

3 . The method of claim 2 , further comprising analyzing, by the risk management server, the third risky transactions based on the acceptable risk threshold selection to identify the suspected fraudulent transactions.

4 . The method of claim 3 , wherein the transaction frequency threshold is based on a rate of cross-border transactions that are fraudulent.

5 . The method of claim 3 , further comprising:

automatically selecting, by the risk management server, the transaction frequency threshold.

6 . The method of claim 3 , wherein the GUI is to further display selectable transaction frequency thresholds, the method further comprising:

receiving, by the risk management server from the issuer server, a selection for the transaction frequency threshold.

7 . The method of claim 3 , wherein the risk management server is a payment processor server.

8 . The method of claim 3 , wherein the risk management server is a payment servicer server.

9 . The method of claim 1 , wherein the fraud transaction parameters comprise a transaction ID, a transaction timestamp, a merchant category code, a transaction type, a merchant name, a sender account number, a transaction amount, a transaction date, transaction time, an issuing country code, or a recipient country code, or a combination thereof.

10 . The method of claim 1 , wherein the machine learning model comprises an ensemble learning model.

11 . The method of claim 1 , wherein the machine learning model comprises a decision tree model.

12 . The method of claim 1 , wherein the machine learning model comprises a random forest model.

13 . The method of claim 1 , wherein the risk management server is a payment processor server.

14 . The method of claim 1 , wherein the risk management server is a payment servicer server.

15 . A computer-implemented method comprising:

applying, by a risk management server, fraudulent transaction data to a machine learning model to determine fraud transaction parameters associated with known fraudulent transactions, and wherein the known fraudulent transactions are associated with merchants;

generating, by the risk management server, a graphical user interface (GUI) for display via an issuer server, the GUI comprising:

risk levels corresponding to the fraud transaction parameters,

threshold fraud rates, and

acceptable risk thresholds;

receiving, by the risk management server from the issuer server via the GUI, a rule selection data signal comprising:

a risk level selection;

a threshold fraud rate selection; and

an acceptable risk threshold selection;

receiving, by the risk management server, a transaction data signal comprising live transaction data associated with live transactions being processed by the issuer server;

identifying, by the risk management server, first risky transactions from the live transactions based on the risk level selection and the fraud transaction parameters;

identifying, by the risk management server, high risk merchants from the merchants based on the threshold fraud rate selection;

identifying, by the risk management server, second risky transactions from the live transactions, wherein the second risky transactions are associated with the high risk merchants;

analyzing, by the risk management server, the first risky transactions and the second risky transactions based on the acceptable risk threshold selection to identify suspected fraudulent transactions;

transmitting, by the risk management server, a transaction authorization signal to the issuer server to cause the first risky transactions and the second risky transactions satisfying the acceptable risk threshold selection to be automatically authorized;

determining, by the risk management server, a fraud detection rate based on indicative of a proportion or a percentage of the known fraudulent transactions identified as suspected fraudulent transactions; and

receiving, by the risk management server from the issuer server via the GUI, an updated rule selection data signal, the updated rule selection data comprising at least one of an updated risk level selection, an updated threshold fraud rate selection, or an updated acceptable risk threshold selection, and the updated rule selection data configured to increase the fraud detection rate.

16 . The method of claim 15 , further comprising:

displaying, by the risk management server via the GUI, the fraud detection rate.

17 . The method of claim 16 , further comprising:

identifying, by the risk management server, high frequency merchants from the merchants, wherein the high frequency merchants are associated with a transaction total that exceeds a transaction frequency threshold; and

identifying, by the risk management server, third risky transactions from the live transactions, wherein the third risky transactions are associated with the high frequency merchants.

18 . The method of claim 17 , further comprising analyzing, by the risk management server, the third risky transactions based on the acceptable risk threshold selection to identify the suspected fraudulent transactions.

19 . A computer-implemented method comprising:

receiving, by a risk management server, transaction parameters for known fraudulent transactions;

identifying, by the risk management server, fraud transaction parameters by applying the transaction parameters to a machine learning model, wherein the fraud transaction parameters comprise ranges of fraud parameter values;

associating, by the risk management server, the ranges of fraud parameter values with risk levels;

receiving, by the risk management server from a transaction data server, test transaction parameters for test transactions corresponding to merchants;

classifying, by the risk management server, the merchants associated with a fraud rate that exceeds a threshold fraud rate as high risk merchants;

transmitting, by the risk management server, a graphical user interface (GUI) signal for accessing a GUI to an issuer server, wherein the GUI is to display:

the risk levels; and

the high risk merchants;

receiving, by the risk management server from the issuer server, a risk level selection data signal based on a risk level selection at the issuer server via the GUI;

determining, by the risk management server, suspected fraudulent test transactions in the test transactions based on the risk level selection;

determining, by the risk management server, a fraud detection rate based on the suspected fraudulent test transactions and the known fraudulent transactions, the fraud detection rate indicative of a proportion or a percentage of the known fraudulent transactions identified as suspected fraudulent test transactions;

receiving, by the risk management server from the issuer server, an updated risk level selection data signal based on an updated risk level selection at the issuer server via the GUI, the updated risk level selection configured to increase the fraud detection rate;

receiving, by the risk management server, a transaction data signal comprising live transaction data associated with live transactions being processed by the issuer server;

analyzing, by the risk management server, the live transactions based on the risk level selection to identify suspected fraudulent live transaction in the live transactions; and

transmitting, by the risk management server, a transaction denial signal to the issuer server to cause the suspected fraudulent live transactions to be automatically denied.