# US 2021/0344635 A1 · Computer-implemented systems configured for automated machine learning contact priority prediction for electronic messages and methods of use thereof

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

- Number: US 2021/0344635 A1
- Document ID: US20210344635A1
- Status: Published application
- Kind: A1
- Application number: 17/323,497
- Filed: 2021-05-18
- Published: 2021-11-04
- Earliest claimed: 2020-04-30
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Adam Vukich; George Bergeron; James Zarakas
- Web page: https://patentgrep.com/patent/US20210344635A1

## CPC

- G06N 3/084: Backpropagation, e.g. using gradient descent
- H04L 51/226: Delivery according to priorities
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/09: Supervised learning
- G06N 20/00: Machine learning
- H04L 51/224: providing notification on incoming messages, e.g. pushed notifications of received messages
- H04L 51/42: Mailbox-related aspects, e.g. synchronisation of mailboxes
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- H04L 51/18: Commands or executable codes
- 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

In order to facilitate automatic message prioritization, systems and methods are described including a processor that receives electronic messages, where each electronic message is associated with a sender and a recipient. The processor utilizes a message prioritization machine learning model to predict a current prioritized ordering of the electronic messages based on parameters associated with each electronic message, where the parameters include a calendar event parameter representing a calendar event associated with each electronic message, a personnel parameter associated with the sender of each electronic message, and a message subject parameter associated with a subject of each electronic message, where the current prioritized ordering includes an order of notification of each electronic message according to priority based on an interaction history of historical electronic messages associated with the sender The processor causes to display the electronic messages according to the current prioritized ordering.

## Claims

### Claim 1

A method comprising:
accessing, by at least one processor, a plurality of electronic messages of an electronic messaging software application associated with a recipient, wherein the plurality of electronic messages are associated with at least one sender and a subject;
accessing, by the at least one processor, a plurality of calendar events residing in an electronic calendar associated with the recipient, wherein the plurality of calendar events are associated with at least one invitee and at least one event subject;
determining, by the at least one processor, at least one calendar event of the plurality of calendar events associated with at least one electronic message of the plurality of electronic messages based at least in part on a matching of:
the at least one sender with the at least one invitee, and
the subject with the at least one event subject;
utilizing, by the at least one processor, a message prioritization machine learning model to predict a current prioritized ordering of the plurality of electronic messages based at least in part on at least one parameter associated with each of the plurality of electronic messages;
wherein the at least one parameter comprises at least one calendar event parameter that represents a calendar event of the at least one calendar event associated with a respective electronic message,
wherein the current prioritized ordering comprises an order of notification of each electronic message of the at least one electronic message according to a message priority of each electronic message at a current time; and
causing to display, by the at least one processor, the plurality of electronic messages according to the current prioritized ordering on a screen of at least one computing device associated with the recipient.

### Claim 2 (depends on 1)

The method of claim 1, wherein the message priority of each respective electronic message of the at least one electronic message is based on an interaction history associated with historical electronic messages associated with each respective sender; and
wherein the interaction history comprises:
i) view data representing a time between receiving and opening each historical electronic message of the historical electronic messages associated with the at least one sender; and
ii) a response data representing a time between opening and transmitting a response to each historical electronic message of the historical electronic messages associated with the at least one sender.

### Claim 3 (depends on 1)

The method of claim 1, wherein the message prioritization machine learning model is trained according to an interaction history to correlate a respective message priority with a respective at least one parameter associated with each respective historical electronic message of the historical electronic messages based on the interaction history associated with each respective historical electronic message.

### Claim 4 (depends on 1)

The method of claim 1, wherein the at least one parameter comprises at least one personnel parameter comprises a hierarchical position data representing a hierarchical position of the at least one sender within an organization based on a hierarchical organization data.

### Claim 5 (depends on 1)

The method of claim 1, further comprising utilizing, by the at least one processor, a text recognition model to determine the at least one message subject parameter of the respective electronic message based on a text string associated with the subject of the respective electronic message.

### Claim 6 (depends on 1)

The method of claim 1, wherein the at least one calendar event parameter of the respective electronic message comprises a temporal proximity of the display of each respective at least one electronic message to the calendar event associated with the respective electronic message.

### Claim 7 (depends on 6)

The method of claim 6, wherein the current prioritized ordering of each respective at least one electronic message dynamically adjusts based on the temporal proximity of each respective calendar event.

### Claim 8 (depends on 1)

The method of claim 1, further comprising utilizing, by the at least one processor, the message prioritization machine learning model to predict an updated prioritized ordering of the at least one electronic message based on one or more user interactions associated with the at least one electronic message.

### Claim 9 (depends on 8)

The method of claim 8, wherein:
the one or more user interactions comprises a read indication associated with one or more electronic messages of the at least one electronic messages; and
the read indication represents that the recipient has read the one or more electronic messages.

### Claim 10 (depends on 8)

The method of claim 8, wherein the updated prioritized ordering is predicted upon each user interaction of the one or more user interactions.

### Claim 11

A method comprising:
receiving, by at least one processor, a plurality of electronic messages of an electronic messaging software application associated with a recipient, wherein the plurality of electronic messages are associated with at least one sender and a subject;
generating, by the at least one processor, a set of message data associated with each of the plurality of electronic messages;
wherein each respective set of message data comprises:
i) at least one sender indicator representing the at least one sender of each respective at least one electronic message, and
ii) at least one message subject indicator representing a respective subject of each respective at least one electronic message;
receiving, by at least one processor, a plurality of calendar events residing in an electronic calendar associated with the recipient, wherein the plurality of calendar events are associated with at least one invitee and at least one event subject;
determining, by the at least one processor, at least one calendar event of the plurality of calendar events associated with at least one electronic message of the plurality of electronic messages based at least in part on a matching of:
the at least one sender with the at least one invitee, and
the subject with the at least one event subject;
transmitting, by the at least one processor, each respective set of message data and the at least one calendar event to a message prioritization system;
wherein the message prioritization system comprises at least one message prioritization processor configured to:
utilize a message prioritization machine learning model to predict a current prioritized ordering of the plurality of electronic messages based at least in part on a at least one parameter associated with each of the plurality of electronic messages;
wherein the at least one parameter comprises at least one calendar event parameter that represents a calendar event of the at least one calendar event associated with a respective electronic message,
wherein the current prioritized ordering comprises an order of notification of each electronic message of the plurality of electronic messages according to a message priority of each electronic message at a current time; and
transmit the current prioritized ordering to the at least one processor; and
causing to display, by the at least one processor, the plurality of electronic messages according to the current prioritized ordering on a screen in communication with the at least one processor.

### Claim 12 (depends on 11)

The method of claim 11, wherein the message priority of each respective electronic message of the at least one electronic message is based on an interaction history associated with historical electronic messages associated with each respective sender; and
wherein the interaction history comprises:
i) view data representing a time between receiving and opening each historical electronic message of the historical electronic messages associated with the at least one sender; and
ii) a response data representing a time between opening and transmitting a response to each historical electronic message of the historical electronic messages associated with the at least one sender.

### Claim 13 (depends on 11)

The method of claim 11, wherein the message prioritization machine learning model is trained according to an interaction history to correlate a respective message priority with a respective at least one parameter associated with each respective historical electronic message of the historical electronic messages based on the interaction history associated with each respective historical electronic message.

### Claim 14 (depends on 11)

The method of claim 11, wherein the at least one parameter comprises at least one personnel parameter comprises a hierarchical position data representing a hierarchical position of the at least one sender within an organization based on a hierarchical organization data.

### Claim 15 (depends on 11)

The method of claim 11, wherein the at least one message prioritization processor is further configured to utilize a text recognition model to determine the at least one message subject parameter of the respective electronic message based on a text string associated with the at least one subject indicator of the respective electronic message.

### Claim 16 (depends on 11)

The method of claim 11, wherein the at least one calendar event parameter of the respective electronic message comprises a temporal proximity of the display of each respective at least one electronic message to the calendar event associated with the respective electronic message.

### Claim 17 (depends on 16)

The method of claim 16, wherein the current prioritized ordering of each respective electronic message of the plurality of electronic messages dynamically adjusts based on the temporal proximity of each respective calendar event.

### Claim 18 (depends on 11)

The method of claim 11, wherein the at least one message prioritization processor is further configured to utilize the message prioritization machine learning model to predict an updated prioritized ordering of the plurality of electronic messages based on one or more user interactions associated with the at least one electronic message.

### Claim 19 (depends on 18)

The method of claim 18, wherein:
the one or more user interactions comprises a read indication associated with one or more electronic messages of the at least one electronic messages; and
the read indication represents that the recipient has read the one or more electronic messages.

### Claim 20

A system comprising:
a non-transitory computer-readable memory having instructions stored thereon; and
at least one processor in communication with the non-transitory computer-readable memory and configured to perform the instructions comprising:
access a plurality of electronic messages of an electronic messaging software application associated with a recipient, wherein the plurality of electronic messages are associated with at least one sender and subject;
access a plurality of calendar events residing in an electronic calendar associated with the recipient, wherein the plurality of calendar events are associated with at least one invitee and at least one event subject;
determine at least one calendar event of the plurality of calendar events associated with at least one electronic message of the plurality of electronic messages based at least in part on a matching of:
the at least one sender with the at least one invitee, and
the subject with the at least one event subject;
utilize a message prioritization machine learning model to predict a current prioritized ordering of the plurality of electronic messages based at least in part on a at least one parameter associated with each of the plurality of electronic messages;
wherein the at least one parameter comprises at least one calendar event parameter that represents a calendar event of the at least one calendar event associated with a respective electronic message,
wherein the current prioritized ordering comprises an order of notification of each electronic message of the at least one electronic message according to a message priority of each electronic message at a current time; and
cause to display the plurality of electronic messages according to the current prioritized ordering on a screen of at least one computing device associated with the recipient.
