Granted Patent B2

US 12,236,439 · App. 18/207,069 · Granted Feb 25, 2025

Inventors: Timothy Kramme (Parker, TX); Elizabeth A. Flowers (Bloomington, IL); Reena Batra (Alpharetta, GA); Miriam Valero (Bloomington, IL); Puneit Dua (Bloomington, IL); Shanna L. Phillips (Bloomington, IL); Russell Ruestman (Minonk, IL); Bradley A. Craig (Normal, IL)

Assignee: State Farm Mutual Automobile Insurance Company

G06Q30/0185G06N5/046G06N20/00G06Q20/102G06Q20/20G06Q20/24G06Q20/3224G06Q20/34G06Q20/401G06Q20/4016G06Q20/407G06Q20/409G06Q30/0225G06V30/194G06V30/41G06Q30/0248

Abstract

Claims (72)

1. A computer-implemented method of using a machine-learned model to determine a cause of a false positive fraud alert, the method comprising:

receiving, by one or more processors, transaction data associated with a financial transaction initiated by a customer;

determining, using a rules engine applying one or more fraud detection rules, a fraud alert associated with the financial transaction;

receiving, by the one or more processors and from a customer computing device, customer feedback indicating that the fraud alert is a false positive fraud alert;

in response to receiving the customer feedback, providing the transaction data as input to a machine-learned model trained to output a piece of data within the transaction data that caused the false positive fraud alert;

determining, based on the output of the machine-learned model, a cause associated with the false positive fraud alert; and

modifying, based on the cause, a first fraud detection rule of the rules engine including a threshold associated with the piece of data that caused the false positive fraud alert.

2. The computer-implemented method of claim 1 , wherein the trained machine-learned model is configured to identify, as the cause associated with the false positive fraud alert, at least one of:

a transaction amount of the financial transaction;

a merchant associated with the financial transaction;

a purchased product or service associated with the financial transaction;

a transaction time of the financial transaction; or

a transaction type associated with the financial transaction.

3. The computer-implemented method of claim 2 , wherein the trained machine-learned model is further configured to identify, as a second cause associated with the false positive fraud alert, the first fraud detection rule of the rules engine.

4. The computer-implemented method of claim 1 , wherein modifying the first fraud detection rule of the rules engine comprises adding a fraud detection rule or subtracting a fraud detection rule from a set of rules applied by the rules engine.

5. The computer-implemented method of claim 1 , wherein modifying the first fraud detection rule comprises modifying at least one of:

a transaction amount criteria of the first fraud detection rule;

a merchant criteria of the first fraud detection rule;

a purchased product or service criteria of the first fraud detection rule; or

a transaction type criteria of the first fraud detection rule.

6. The computer-implemented method of claim 1 , wherein modifying the first fraud detection rule comprises:

providing the cause associated with the false positive fraud alert to a second machine-learned model different from the trained machine-learned model; and

determining, based at least in part on an output of the second machine-learned model, a modification to the first fraud detection rule.

7. The computer-implemented method of claim 1 , wherein receiving the customer feedback comprises:

based on determining the fraud alert, transmitting an electronic fraud alert to the customer computing device; and

receiving the customer feedback from the customer computing device, responsive to the electronic fraud alert, indicating that the fraud alert is a false positive.

8. The computer-implemented method of claim 1 , wherein identifying the piece of data that caused the false positive fraud alert comprises identifying one or more inconsistencies between a fact pattern associated with the financial transaction, and a historical transaction usage pattern of the customer.

9. The computer-implemented method of claim 1 , wherein the trained machine-learned model is trained to output a fact pattern associated with the false positive fraud alert, wherein the fact pattern includes a combination of two or more of:

a card type;

a card issuer;

a card number;

a cardholder name;

a merchant name;

a merchant location;

a transaction location;

a transaction amount; and

a transaction type.

10. The computer-implemented method of claim 9 , further comprising:

receiving second transaction data associated with a second financial transaction;

determining, using the rules engine, that the second financial transaction corresponds to the fact pattern associated with the false positive fraud alert; and

causing, by the rules engine, based at least in part on the determining that the second financial transaction corresponds to the fact pattern, at least one of:

deferring transmission of an electronic fraud alert associated with the second financial transaction; or

deferring an account freeze associated with the second financial transaction.

11. The computer-implemented method of claim 1 , further comprising:

causing at least one of an electronic fraud alert or an account freeze associated with the financial transaction not to be performed, based on receiving the indication that the fraud alert is a false positive.

12. A system for using a trained machine-learned model to determine a cause of a false positive fraud alert, the system comprising:

one or more processors; and

one or more non-transitory computer-readable media storing computer-executable instructions that, when executed, cause the one or more processors to perform operations comprising:

receiving transaction data associated with a financial transaction initiated by a customer;

determining, using a rules engine applying one or more fraud detection rules, a fraud alert associated with the financial transaction;

receiving, from a customer computing device, customer feedback indicating that the fraud alert is a false positive fraud alert;

in response to receiving the customer feedback, providing the transaction data as input to a machine-learned model trained to output a piece of data within the transaction data that caused the false positive fraud alert; that provides an output;

determining, based on the output of the machine-learned model, a cause associated with the false positive fraud alert; and

modifying, based on the cause, a first fraud detection rule of the rules engine, wherein modifying the first fraud detection rule includes changing including a threshold associated with the piece of data that caused the false positive fraud alert.

13. The system of claim 12 , wherein the trained machine-learned model is configured to identify, as the cause associated with the false positive fraud alert, at least one of:

a transaction amount of the financial transaction;

a merchant associated with the financial transaction;

a purchased product or service associated with the financial transaction;

a transaction time of the financial transaction; or

a transaction type associated with the financial transaction.

14. The system of claim 13 , wherein the trained machine-learned model is further configured to identify, as a second cause associated with the false positive fraud alert, the first fraud detection rule of the rules engine.

15. The system of claim 12 , wherein modifying the first fraud detection rule of the rules engine comprises adding a fraud detection rule or subtracting a fraud detection rule from a set of rules applied by the rules engine.

16. The system of claim 12 , wherein modifying the first fraud detection rule comprises modifying at least one of:

a transaction amount criteria of the first fraud detection rule;

a merchant criteria of the first fraud detection rule;

a purchased product or service criteria of the first fraud detection rule; or

a transaction type criteria of the first fraud detection rule.

17. The system of claim 12 , wherein receiving the customer feedback comprises:

based on determining the fraud alert, transmitting an electronic fraud alert to the customer computing device; and

receiving the customer feedback from the customer computing device, responsive to the electronic fraud alert, indicating that the fraud alert is a false positive.

18. The system of claim 12 , the operations further comprising:

causing at least one of an electronic fraud alert or an account freeze associated with the financial transaction not to be performed, based on receiving the indication that the fraud alert is a false positive.

Assignments (1)

Continuity (7)

References Cited (400)