Updating a machine learning fraud model based on third party transaction information

Patent metadata

Field	Value
Publication number	US-11954689-B2
Application number	US-202218063914-A
Country	US
Kind code	B2
Filing date	Dec 9, 2022
Priority date	Feb 19, 2019
Publication date	Apr 9, 2024
Grant date	Apr 9, 2024

Abstract

A device receives first transaction information associated with a first transaction, and a first transaction account utilized for the first transaction and associated with a first financial institution. The device determines, based on a fraud model, that the first transaction is to be denied due to potential fraud associated with the first transaction account and receives second transaction information associated with a second transaction, and a second transaction account utilized for the second transaction and associated with a second financial institution. The device processes the first transaction information and the second transaction information, with a matching model, to determine whether the first transaction information matches the second transaction information and determines that the first transaction was incorrectly denied when the first transaction information matches the second transaction information within a predetermined threshold. The device performs one or more actions based on determining that the first transaction was incorrectly denied.

First claim

What is claimed is:

1. A method, comprising:
receiving, by a device, first information associated with a first transaction that is related to a first account associated with a first entity,
wherein the first transaction is associated with utilizing a first transaction card;

determining, by the device and based on a first machine learning model, that the first transaction is to be denied due to a potential fraud associated with the first account;
obtaining, by the device, second information associated with a second transaction that is related to a second account associated with a second entity,
wherein the second transaction is associated with utilizing a second transaction card, and
wherein the first entity and the second entity are different;

determining, by the device and based on a second machine learning model, whether the first information and the second information match within a predetermined threshold; and
performing, by the device and based on determining whether the first information and the second information match within the predetermined threshold, an action that includes:
removing, when the first information and the second information match within the predetermined threshold, a fraud lock associated with the first account to allow the first account to be utilized, or
determining that the first transaction was correctly denied due to the potential fraud.

2. The method of claim 1 , wherein the second information is obtained based on scraping the second information from an application executing on another device.

3. The method of claim 1 , wherein the second information is received from third party sources that are not associated with the second entity.

4. The method of claim 1 , wherein the second information is received from another device associated with the second account.

5. The method of claim 1 , wherein the first account and the second account are associated with a user.

6. The method of claim 1 , wherein the first information and the second information are associated with at least one of:
a first amount associated with the first transaction and a second amount associated with the second transaction,
first merchant information associated with the first transaction and second merchant information associated with the second transaction,
first location information associated with the first transaction and second location information associated with the second transaction,
first information indicating whether the first transaction occurred online or at a physical location, and second information indicating whether the second transaction occurred online or at a physical location, or
first time information associated with the first transaction and second time information associated with the second transaction.

7. The method of claim 1 , further comprising:
updating training data for the first machine learning model, based on parameters of the first machine learning model that caused the first transaction to be incorrectly denied.

8. A device, comprising:
one or more memories; and
one or more processors, coupled to the one or more memories, configured to:
receive first information associated with a first transaction that is related to a first account associated with a first entity,
wherein the first transaction is associated with utilizing a first transaction card;

determine, based on a first machine learning model, that the first transaction is to be denied due to a potential fraud associated with the first account;
obtain second information associated with a second transaction that is related to a second account associated with a second entity,
wherein the second transaction is associated with utilizing a second transaction card;

determine, based on a second machine learning model, whether the first information and the second information match within a predetermined threshold; and
perform, based on determining whether the first information and the second information match within the predetermined threshold, an action that includes:
removing, when the first information and the second information match within the predetermined threshold, a fraud lock associated with the first account to allow the first account to be utilized, or
determining that the first transaction was correctly denied due to the potential fraud.

9. The device of claim 8 , wherein the second information is obtained based on scraping the second information from an application executing on another device.

10. The device of claim 8 , wherein the second information is received from third party sources that are not associated with the second entity.

11. The device of claim 8 , wherein the second information is received from another device associated with the second account.

12. The device of claim 8 , wherein the first account and the second account are associated with a user.

13. The device of claim 8 , wherein the first information and the second information are associated with at least one of:
a first amount associated with the first transaction and a second amount associated with the second transaction,
first merchant information associated with the first transaction and second merchant information associated with the second transaction,
first location information associated with the first transaction and second location information associated with the second transaction,
first information indicating whether the first transaction occurred online or at a physical location, and second information indicating whether the second transaction occurred online or at a physical location, or
first time information associated with the first transaction and second time information associated with the second transaction.

14. The device of claim 8 , wherein the one or more processors are further configured to:
update training data for the first machine learning model, based on parameters of the first machine learning model that caused the first transaction to be incorrectly denied.

15. A non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the device to:
receive first information associated with a first transaction that is related to a first account associated with a first entity;
determine, based on a first machine learning model, that the first transaction is to be denied due to a potential fraud associated with the first account;
obtain second information associated with a second transaction that is related to a second account associated with a second entity,
wherein the first entity and the second entity are different; and wherein the first and second transactions are associated with utilizing different transaction cards;

determine, based on a second machine learning model, whether the first information and the second information match within a predetermined threshold; and
perform, based on determining whether the first information and the second information match within the predetermined threshold, an action that includes:
removing, when the first information and the second information match within the predetermined threshold, a fraud lock associated with the first account to allow the first account to be utilized, or
determining that the first transaction was correctly denied due to the potential fraud.

16. The non-transitory computer-readable medium of claim 15 , wherein the second information is obtained based on scraping the second information from an application executing on another device.

17

Assignees

• Capital One Services Llc

Inventors

• Hart Colin
• Edwards Joshua
• Perezleon Francisco
• Johnson Molly
• Newman Kaitlin
• Wu Angelina
• Ji Jason

Classifications

• H04L63/20

  for managing network security; network security policies in general (filtering policies H04L63/0227) · CPC title

• H04L63/102

  Entity profiles · CPC title

• G06Q20/4016 Primary

  involving fraud or risk level assessment in transaction processing · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06Q20/401

  Transaction verification · CPC title