# US 11,907,997 B2 · Device, method, and computer-readable media for recommendation networking based on connections, characteristics, and assets using machine learning

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

- Number: US 11,907,997 B2
- Document ID: US11907997B2
- Status: Granted
- Kind: B2
- Application number: 17/488,505
- Filed: 2021-09-29
- Published: 2023-03-30
- Granted: 2024-02-20
- Earliest claimed: 2021-09-29
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Micah Price; Avid Ghamsari; Qiaochu Tang; Geoffrey Dagley; Staeven Duckworth
- Web page: https://patentgrep.com/patent/US11907997B2

## CPC

- G06Q 30/0631: {Recommending goods or services}
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06Q 30/0625: {by formulating product or service queries, e.g. using keywords or predefined options}
- H04L 51/52: for supporting social networking services
- H04L 51/18: Commands or executable codes
- 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
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Methods, systems, and apparatuses are described herein for providing purchase recommendations by analyzing social networks using machine learning. A machine learning model may be trained to select one or more of the first plurality of users. Purchase intention data that indicates an intention of a first user to acquire a type of asset may be received. Social networking data that comprises a plurality of associations between a second plurality of users may be received. Purchase history data indicating one or more purchases, of one or more assets associated with the type of asset, made by the second plurality of users may be received. The trained machine learning model may be provided the data. In return, the trained machine learning model may provide an indication of a second user. A notification may be sent to the second user.

## Claims

### Claim 1

A computing device comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the computing device to:
determine, for a demographic of users, training data indicating:
first connections between a first plurality of users that are members of the demographic,
first characteristics of the first plurality of users that are members of the demographic,
a first plurality of assets purchased by the first plurality of users that are members of the demographic, and
a history of communications between the first plurality of users relating to the first plurality of assets;
train, using the training data, a machine learning model, implemented via an artificial neural network and comprising a plurality of nodes, to:
receive input indicating second connections between a third plurality of users, second characteristics of the third plurality of users, and a second plurality of assets purchased by the third plurality of users, and
output, based on the input, one or more of the third plurality of users, wherein training the machine learning model comprises modifying, based on the training data, a weight between at least two of the plurality of nodes such that, when provided at least a portion of the training data, the machine learning model outputs one of the first plurality of users based on the first connections, the first characteristics, and the first plurality of assets;
determine purchase intention data that indicates an intention of a first user to acquire a type of asset by collecting web activity data by monitoring web browsing activity of the first user, wherein the purchase intention data indicates:
characteristics of the first user, and
preferences of the first user corresponding to the type of asset;
determine social networking data that comprises a plurality of associations between a second plurality of users;
determine purchase history data indicating one or more purchases, of one or more assets associated with the type of asset, made by the second plurality of users;
process the purchase history data to identify one or more merchants where the second plurality of users purchased the one or more assets;
input, to an input node of the plurality of nodes of the trained machine learning model, input data comprising:
the purchase intention data,
an indication of the one or more merchants,
the social networking data, and
the purchase history data;
determine output from an output node of the plurality of nodes of the trained machine learning model that comprises at least one second user of the second plurality of users, wherein the at least one second user has purchased a second asset that is the type of asset;
generate, for the at least one second user, a notification prompting the second user to contact the first user regarding the intention of the first user to acquire the type of asset; and
cause the notification to be transmitted to the second user.

### Claim 2 (depends on 1)

The computing device of claim 1, wherein the instructions, when executed by the one or more processors, cause the computing device to determine the purchase intention data by causing the computing device to:
detect, based on the web activity data, that the first user is browsing web pages associated with the type of asset.

### Claim 3 (depends on 1)

The computing device of claim 1, wherein the instructions, when executed by the one or more processors, cause the computing device to:
determine a similarity between the type of asset and the one or more assets, wherein the input data further comprises an indication of the similarity between the type of asset and the one or more assets.

### Claim 4 (depends on 1)

The computing device of claim 1, wherein the instructions, when executed by the one or more processors, cause the computing device to:
determine a similarity between the characteristics of the first user and characteristics of the second plurality of users, wherein the input data further comprises an indication of the similarity between the characteristics of the first user and characteristics of the one or more users.

### Claim 5 (depends on 1)

The computing device of claim 1, wherein the input data further comprises an indication of whether the first user and the second plurality of users are members of a group.

### Claim 6 (depends on 1)

The computing device of claim 1, wherein the purchase history data indicates a recency of purchase, one or more assets, by the second plurality of users.

### Claim 7 (depends on 1)

The computing device of claim 1, wherein at least a portion of the first plurality of users is the same as the second plurality of users.

### Claim 8 (depends on 1)

The computing device of claim 1, wherein the history of communications between the first plurality of users relating to the first plurality of assets comprises one or more indications of properties of the first plurality of assets.

### Claim 9

A method comprising:
determining, for a demographic of users, training data indicating:
first connections between a first plurality of users that are members of the demographic,
first characteristics of the first plurality of users that are members of the demographic,
a first plurality of assets purchased by the first plurality of users that are members of the demographic, and
a history of communications between the first plurality of users relating to the first plurality of assets;
training, using the training data, a machine learning model, implemented via an artificial neural network and comprising a plurality of nodes, to:
receive input indicating second connections between a third plurality of users, second characteristics of the third plurality of users, and a second plurality of assets purchased by the third plurality of users, and
output, based on the input, one or more of the third plurality of users, wherein training the machine learning model comprises modifying, based on the training data, a weight between at least two of the plurality of nodes such that, when provided at least a portion of the training data, the machine learning model outputs one of the first plurality of users based on the first connections, the first characteristics, and the first plurality of assets;
determining purchase intention data that indicates an intention of a first user to acquire a type of asset by collecting web activity data by monitoring web browsing activity of the first user, wherein the purchase intention data indicates:
characteristics of the first user, and
preferences of the first user corresponding to the type of asset;
determining social networking data that comprises a plurality of associations between a second plurality of users;
determining purchase history data indicating one or more purchases, of one or more assets associated with the type of asset, made by the second plurality of users;
processing the purchase history data to identify one or more merchants where the second plurality of users purchased the one or more assets;
inputting, to an input node of the plurality of nodes of the trained machine learning model, input data comprising:
the purchase intention data,
an indication of the one or more merchants,
the social networking data, and
the purchase history data;
determining output from an output node of the plurality of nodes of the trained machine learning model that comprises at least one second user of the second plurality of users, wherein the at least one second user has purchased a second asset that is the type of asset;
generating, for the at least one second user, a notification prompting the second user to contact the first user regarding the intention of the first user to acquire the type of asset; and
causing the notification to be transmitted to the second user.

### Claim 10 (depends on 9)

The method of claim 9, wherein determining the purchase intention data comprises:
detecting, based on the web activity data, that the first user is browsing web pages associated with the type of asset.

### Claim 11 (depends on 9)

The method of claim 9, further comprising:
determining a similarity between the type of asset and the one or more assets, wherein the input data further comprises an indication of the similarity between the type of asset and the one or more assets.

### Claim 12 (depends on 9)

The method of claim 9, further comprising:
determining a similarity between the characteristics of the first user and characteristics of the second plurality of users, wherein the input data further comprises an indication of the similarity between the characteristics of the first user and characteristics of the one or more users.

### Claim 13 (depends on 9)

The method of claim 9, wherein the input data further comprises an indication of whether the first user and the second plurality of users are members of a group.

### Claim 14 (depends on 9)

The method of claim 9, wherein the history of communications between the first plurality of users relating to the first plurality of assets comprises one or more indications of properties of the first plurality of assets.

### Claim 15

One or more non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause a computing device to:
determine, for a demographic of users, training data indicating:
first connections between a first plurality of users that are members of the demographic,
first characteristics of the first plurality of users that are members of the demographic,
a first plurality of assets purchased by the first plurality of users that are members of the demographic, and
a history of communications between the first plurality of users relating to the first plurality of assets;
train, using the training data, a machine learning model, implemented via an artificial neural network and comprising a plurality of nodes, to:
receive input indicating second connections between a third plurality of users, second characteristics of the third plurality of users, and a second plurality of assets purchased by the third plurality of users, and
output, based on the input, one or more of the third plurality of users, wherein training the machine learning model comprises modifying, based on the training data, a weight between at least two of the plurality of nodes such that, when provided at least a portion of the training data, the machine learning model outputs one of the first plurality of users based on the first connections, the first characteristics, and the first plurality of assets;
determine purchase intention data that indicates an intention of a first user to acquire a type of asset by collecting web activity data by monitoring web browsing activity of the first user, wherein the purchase intention data indicates:
characteristics of the first user, and
preferences of the first user corresponding to the type of asset;
determine social networking data that comprises a plurality of associations between a second plurality of users;
determine purchase history data indicating one or more purchases, of one or more assets associated with the type of asset, made by the second plurality of users;
process the purchase history data to identify one or more merchants where the second plurality of users purchased the one or more assets;
input, to an input node of the plurality of nodes of the trained machine learning model, input data comprising:
the purchase intention data,
an indication of the one or more merchants,
the social networking data, and
the purchase history data;
determine output from an output node of the plurality of nodes of the trained machine learning model that comprises at least one second user of the second plurality of users, wherein the at least one second user has purchased a second asset that is the type of asset;
generate, for the at least one second user, a notification prompting the second user to contact the first user regarding the intention of the first user to acquire the type of asset; and
cause the notification to be transmitted to the second user.

### Claim 16 (depends on 15)

The non-transitory computer-readable media of claim 15, wherein the instructions, when executed by the one or more processors, cause the computing device to determine the purchase intention data by causing the computing device to:
detect, based on the web activity data, that the first user is browsing web pages associated with the type of asset.

### Claim 17 (depends on 15)

The non-transitory computer-readable media of claim 15, wherein the instructions, when executed by the one or more processors, cause the computing device to:
determine a similarity between the type of asset and the one or more assets, wherein the input data further comprises an indication of the similarity between the type of asset and the one or more assets.

### Claim 18 (depends on 15)

The non-transitory computer-readable media of claim 15, wherein the instructions, when executed by the one or more processors, cause the computing device to:
determine a similarity between the characteristics of the first user and characteristics of the second plurality of users, wherein the input data further comprises an indication of the similarity between the characteristics of the first user and characteristics of the one or more users.

### Claim 19 (depends on 15)

The non-transitory computer-readable media of claim 15, wherein the input data further comprises an indication of whether the first user and the second plurality of users are members of a group.

### Claim 20 (depends on 15)

The one or more non-transitory computer-readable media of claim 15, wherein the history of communications between the first plurality of users relating to the first plurality of assets comprises one or more indications of properties of the first plurality of assets.