# US 2025/0165978 A1 · Merchant specific machine learning model for fraud detection

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0165978 A1
- Document ID: US20250165978A1
- Status: Published application
- Kind: A1
- Application number: 18/513,104
- Filed: 2023-11-17
- Published: 2025-05-22
- Earliest claimed: 2023-11-17
- Assignees: Stripe, Inc.
- Applicants: Stripe, Inc.
- Inventors: Logan Howard; Ryan Drapeau; Nathan Brasher
- Web page: https://patentgrep.com/patent/US20250165978A1

## CPC

- G06Q 20/4016: {involving fraud or risk level assessment in transaction processing}
- G06N 20/00: Machine learning
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A server computer system for detecting a fraudulent electronic transaction may be configured to receive a transaction associated with a merchant. The system processes the transaction with a first machine learning model that is trained specific to the merchant, to obtain a first fraud detection, and processes the transaction with a second machine learning model that is trained based on a plurality of merchants, to obtain a second fraud detection. The system detects fraud associated with the transaction based on the first fraud detection and the second fraud detection. In response to detecting the fraud that is associated with the transaction, the system blocks the transaction.

## Claims

### Claim 1

A method performed by a server computer system for detecting a fraudulent electronic transaction, comprising:
receiving, by the server computer system, a transaction associated with a merchant;
processing the transaction with a first machine learning model that is trained specific to the merchant, to obtain a first fraud detection;
processing the transaction with a second machine learning model that is trained based on a plurality of merchants, to obtain a second fraud detection;
detecting fraud associated with the transaction based on the first fraud detection and the second fraud detection; and
blocking the transaction, in response to detecting the fraud that is associated with the transaction.

### Claim 2 (depends on 1)

The method of claim 1, wherein the fraud is detected in response to the first fraud detection being positive or the second fraud detection being positive.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
determining that a quantity of transactions that are blocked by the first machine learning model exceeds a threshold; and
in response to the quantity of transactions exceeding the threshold, disabling blocking of a future transaction associated with the merchant.

### Claim 4 (depends on 1)

The method of claim 1, wherein first training data used to train the first machine learning model comprises first past transactions and is based on a single merchant which is the same as the merchant that the first machine learning model is applied with respect to.

### Claim 5 (depends on 4)

The method of claim 4, wherein second training data used to train the second machine learning model comprises second past transactions from the plurality of merchants.

### Claim 6 (depends on 5)

The method of claim 5, wherein the first past transactions span a shorter period of time than the second past transactions and the first past transactions spans a more recent period of time than the second past transactions.

### Claim 7 (depends on 1)

The method of claim 1, wherein processing the transaction with the first machine learning model comprises matching the merchant to the first machine learning model, wherein the first machine learning model is one of a plurality machine learning models that are each trained specific to a respective merchant.

### Claim 8 (depends on 1)

The method of claim 1, further comprising, without human input, monitoring a plurality of transactions associated with the merchant, and training the first machine learning model in response to fraud activity associated with the plurality of transactions of the merchant satisfying a threshold.

### Claim 9

A non-transitory computer readable storage medium storing instructions, which when executed by a computer processing system, causes the computer processing system to perform operations for detecting a fraudulent electronic transaction, the operations comprising:
receiving, by the computer processing system, a transaction associated with a merchant;
processing the transaction with a first machine learning model that is trained specific to the merchant, to obtain a first fraud detection;
processing the transaction with a second machine learning model that is trained based on a plurality of merchants, to obtain a second fraud detection;
detecting fraud associated with the transaction based on the first fraud detection and the second fraud detection; and
blocking the transaction, in response to detecting the fraud that is associated with the transaction.

### Claim 10 (depends on 9)

The non-transitory computer readable storage medium of claim 9, wherein the fraud is detected in response to the first fraud detection being positive or the second fraud detection being positive.

### Claim 11 (depends on 9)

The non-transitory computer readable storage medium of claim 9, wherein the operations further comprise:
determining that a quantity of transactions that are blocked by the first machine learning model exceeds a threshold; and
in response to the quantity of transactions exceeding the threshold, disabling blocking of a future transaction associated with the merchant.

### Claim 12 (depends on 9)

The non-transitory computer readable storage medium of claim 9, wherein first training data used to train the first machine learning model comprises first past transactions and is performed based on a single merchant which is the same as the merchant that the first machine learning model is applied with respect to.

### Claim 13 (depends on 12)

The non-transitory computer readable storage medium of claim 12, wherein second training data used to train the second machine learning model comprises second past transactions from the plurality of merchants.

### Claim 14 (depends on 9)

The non-transitory computer readable storage medium of claim 9, wherein processing the transaction with the first machine learning model comprises matching the merchant to the first machine learning model, wherein the first machine learning model is one of a plurality machine learning models that are each trained specific to a respective merchant.

### Claim 15

A server computer system for detecting a fraudulent electronic transaction, the server computer system comprising:
a memory; and
one or more processors coupled with the memory configured to perform operations, comprising:
receiving, by the server computer system, a transaction associated with a merchant;
processing the transaction with a first machine learning model that is trained specific to the merchant, to obtain a first fraud detection;
processing the transaction with a second machine learning model that is trained based on a plurality of merchants, to obtain a second fraud detection;
detecting fraud associated with the transaction based on the first fraud detection and the second fraud detection; and
blocking the transaction, in response to detecting the fraud that is associated with the transaction.

### Claim 16 (depends on 15)

The server computer system of claim 15, wherein the fraud is detected in response to the first fraud detection being positive or the second fraud detection being positive.

### Claim 17 (depends on 15)

The server computer system of claim 15, wherein the operations further comprise:
determining that a quantity of transactions that are blocked by the first machine learning model exceeds a threshold; and
in response to the quantity of transactions exceeding the threshold, disabling blocking of a future transaction associated with the merchant.

### Claim 18 (depends on 17)

The server computer system of claim 17, wherein first training data used to train the first machine learning model comprises first past transactions and is based on a single merchant which is the same as the merchant that the first machine learning model is applied with respect to.

### Claim 19 (depends on 18)

The server computer system of claim 18, wherein second training data used to train the second machine learning model comprises second past transactions from the plurality of merchants.

### Claim 20 (depends on 15)

The server computer system of claim 15, wherein processing the transaction with the first machine learning model comprises matching the merchant to the first machine learning model, wherein the first machine learning model is one of a plurality machine learning models that are each trained specific to a respective merchant.
