# US 2025/0225523 A1 · Machine learning engine using following link selection

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

- Number: US 2025/0225523 A1
- Document ID: US20250225523A1
- Status: Published application
- Kind: A1
- Application number: 18/978,965
- Filed: 2024-12-12
- Published: 2025-07-10
- Earliest claimed: 2017-12-28
- Assignees: Paypal, Inc.
- Applicants: PayPal, Inc.
- Inventors: Chern Jie Lim; Ziyuan Pan; Jessica Tjong; Oscar Charles Edward Sanderson; Yanfei Dong
- Web page: https://patentgrep.com/patent/US20250225523A1

## CPC

- G06Q 20/4016: {involving fraud or risk level assessment in transaction processing}
- G06F 16/9566: {URL specific, e.g. using aliases, detecting broken or misspelled links}
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G06Q 20/045: {using payment protocols involving tickets}
- G06Q 20/351: {Virtual cards}
- H04L 63/1425: {Traffic logging, e.g. anomaly detection}
- H04L 63/1483: {service impersonation, e.g. phishing, pharming or web spoofing (detection of rogue wireless access points H04W12/12)}
- G06F 21/31: User authentication
- 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
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A machine learning engine may be trained using artificial intelligence techniques and used according to techniques discussed herein. While an initial electronic transaction for a resource may be permitted, a subsequent related transaction to the initial electronic transaction may be analyzed in view of additional electronic information that was not available at the time of the initial transaction. Analysis of the subsequent related transaction, using the machine learning engine, may indicate a new classification related to the resource and/or the acquisition of the resource. Based on this new classification, usage of the resource may be restricted and/or denied, and the initial transaction for the resource may even be canceled retroactively.

## Claims

### Claim 1

(canceled)

### Claim 2

A system comprising:
a processor; and
a non-transitory computer-readable medium having stored thereon instructions that, when executed by the processor, cause the processor to perform operations comprising:
receiving, from a first user device of a first user via an electronically presented user interface (UI) element in a first UI presented by the system, a request to claim a stored value digital resource using a unique digital identifier, wherein the request is received with a first device identifier of the first user device based on a navigation event associated with the electronically presented UI element;
determining online data associated with at least one of the first device identifier or a first account claiming the stored value digital resource based on the first device identifier, wherein the online data is associated with whether the at least one of the first device identifier or the first account has an indication for fraud;
querying, using the unique digital identifier, a database for account information corresponding to a second account of a second user that established the stored valued digital resource, wherein the account information comprises first information indicating a time of a creation of the stored value digital resource using the second account, and wherein the account information further comprises second information indicating whether the second account has a reversed transaction after the time of the creation of the stored value digital resource;
executing a machine learning engine based on data for a plurality of classification features extracted from at least a portion of the online data and the account information corresponding to the second account;
classifying, based on one or more outputs from executing the machine learning engine, whether the request is a fraudulent request and whether the creation was a fraudulent creation; and
based on the classifying:
allowing the first account to claim the stored value digital resource by transmitting data to the first user device that enables the first user device to claim the stored value digital resource through a second UI presented by the system, or
disallowing the first account from claiming the stored value digital resource by restricting an access to the stored value digital resource through a third UI presented by the system.

### Claim 3 (depends on 2)

The system of claim 2, wherein the operations further comprise:
querying, using the first device identifier, the database for account information corresponding to the first account claiming the stored value digital resource,
wherein the data for the plurality of classification features is further extracted from at least a portion of the account information corresponding to the first account.

### Claim 4 (depends on 3)

The system of claim 3, wherein the account information of at least one of the first account or the second account includes at least one of one or more past IP addresses used by the first account or the second account, one or more gift cards associated with the first account or the second account, or one or more geo-locations associated with the first account or the second account.

### Claim 5 (depends on 2)

The system of claim 2, wherein the stored value digital resource is associated with a gift card having a card recipient, and wherein the plurality of classification features in at least one feature associated with whether the card recipient matches an account identifier associated with the account information corresponding to the second account.

### Claim 6 (depends on 2)

The system of claim 2, wherein the online data comprises at least one of a session IP address associated with a current session for the first user device with the first UI, a browser type and a browser version of a web browser used for the current session, or a browser input to the web browser.

### Claim 7 (depends on 2)

The system of claim 2, wherein the operations further comprise:
executing an action to move the stored value digital resource to an account or secure the unique digital identifier from a further use by the first user device or the first account based on the classifying.

### Claim 8 (depends on 2)

The system of claim 2, wherein the machine learning engine comprises a plurality of decision trees, and wherein the plurality of decision trees are trained using training data associated with at least one of prior claims to previous stored value digital resources or prior uses of the previous stored value digital resources during transaction processing.

### Claim 9

A method comprising:
receiving a unique digital identifier for a stored value digital resource with a request to use the stored value digital resources by a first user via first user interface (UI) provided by a computer system that enables using at least the stored value digital resource, wherein the first UI is presented based on a navigation event via a link selection from the first UI;
determining whether the first user has an indication of fraud;
determining a time of a creation of the stored value digital resource by a second user using a funding instrument and whether the funding instrument is associated with a reversed transaction after the time of the creation of the stored value digital resource;
executing a machine learning engine based on a plurality of classification features associated with whether the first user has an indication of fraud or whether the funding instrument is associated with a reversed transaction;
classifying, based on one or more outputs from executing the machine learning engine, whether the stored value digital resource was fraudulently created or the first user is acting fraudulently by claiming the stored value digital resource; and
executing an action to permit or deny using the stored value digital resource by the first user via a second UI of the computer system based on the classifying.

### Claim 10 (depends on 9)

The method of claim 9, wherein the time of the creation of the stored value digital resource is determined based on account information for one of the funding instrument or an account associated with the second user.

### Claim 11 (depends on 9)

The method of claim 9, wherein the executing the action comprises one of:
allowing the first user to claim the stored value digital resource using an account; or
disallowing the first user from claiming the stored value digital resource and reporting one of the first user or the unique digital identifier as fraudulent.

### Claim 12 (depends on 9)

The method of claim 9, further comprising:
determining account information for an account used by the first user,
wherein the determining whether the first user has the indication of fraud is based on the account information.

### Claim 13 (depends on 12)

The method of claim 12, wherein the account information includes at least one of one or more past IP addresses used by the account, one or more gift cards associated with the account, or one or more geo-locations associated with the account.

### Claim 14 (depends on 9)

The method of claim 9, wherein at least one of the plurality of classification features are associated with gift card redemptions of gift cards corresponding to stored value digital resources.

### Claim 15 (depends on 9)

The method of claim 9, wherein the determining whether the first user has an indication of fraud is based on at least one of a session IP address, browser information, or a browser input associated with a user device of the first user used to access the first UI.

### Claim 16 (depends on 9)

The method of claim 9, wherein the executing the action comprises moving the stored value digital resource to an account of one of the first user or the second user.

### Claim 17

A non-transitory computer-readable medium having stored thereon instructions that, when executed by a processor of a computer system, cause the processor of the computer system to perform operations comprising:
detecting an indication of a request to use a stored digital value by a first user device of a first user via a user interface (UI) element electronically presented in a first UI of an application of an online transaction processor that facilitates uses of stored digital values, wherein the UI element is associated with a navigation link to the application, and wherein the indication is associated with a selection of the navigation link via the UI element that causes the first user device to be directed to the first UI in the application;
determining fraud data associated with the first user based on a first user device identifier of the first user device;
determining historical data associated with the stored digital value that indicates a time of a creation of the stored digital value and whether an account of a second user that created the stored digital value has a reversed transaction at or after the time of the creation of the stored digital value;
processing the fraud data and the historical data using a machine learning engine, wherein the processing comprises:
determining data for input features of the machine learning engine based on the fraud data and the historical data,
executing the machine learning engine based on the data, the input features, and a machine learning process of the machine learning engine, and
outputting a fraud assessment based on the executing; and
based on the fraud assessment, redirecting the first UI to one of a second UI or a third UI based on the request, wherein the second UI facilitates redeeming the stored digital value responsive to the request, and wherein the third UI facilitates preventing access to the stored digital value responsive to the request.

### Claim 18 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the fraud data is associated with whether an account of the first user has been used for fraud sine the time of the creation of the stored digital value.

### Claim 19 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the operations further comprise:
executing an action to move the stored value digital resource to one or more accounts based on the fraud assessment.

### Claim 20 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the historical data comprises account data for the account of the second user.

### Claim 21 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the executing the machine learning engine comprises classifying the input features for the fraud assessment based on the machine learning process.
