Applying rules to improve nearest neighbor matching in a shared embedding space

Patent metadata

Field	Value
Publication number	US-2025131292-A1
Application number	US-202318493495-A
Country	US
Kind code	A1
Filing date	Oct 24, 2023
Priority date	Oct 24, 2023
Publication date	Apr 24, 2025
Grant date	—

Abstract

In some implementations, a machine learning system may receive a set of first data structures associated with a set of users, may receive a set of second data structures associated with a set of entities, and may receive event information associated with the set of users and the set of entities. The machine learning system may generate an embedding space that represents the set of first data structures, the set of second data structures, and the event information. The machine learning system may disregard a portion of the set of second data structures, using a condition, in order to generate a subset of second data structures. The machine learning system may identify at least one relevant entity, from the subset of second data structures, using the embedding space. The machine learning system may generate a communication, associated with the at least one relevant entity, and may output the communication.

First claim

What is claimed is:

1 . A system for applying rules to improve nearest neighbor matching, the system comprising:
one or more memories; and one or more processors, communicatively coupled to the one or more memories, configured to:
receive a set of first data structures associated with a set of users;
receive a set of second data structures associated with a set of entities;
receive event information associated with the set of users and the set of entities;
generate an embedding space that encodes:
a set of first nodes representing the set of users and based on the set of first data structures,
a set of second nodes representing the set of entities and based on the set of second data structures, and
a set of connections, between the set of first nodes and the set of second nodes, based on the event information;

identify a portion of the set of second nodes, for a selected first node in the set of first nodes, using a condition applied to the set of second nodes;
calculate weighted distances using a portion of the set of connections that corresponds to the portion of the set of second nodes and the selected first node;
generate at least one communication based on the weighted distances; and
transmit the at least one communication to a user device associated with a user, in the set of users, represented by the selected first node.

2 . The system of claim 1 , wherein the one or more processors are configured to:
receive a set of credentials associated with the event information, wherein the event information is received using the set of credentials.

3 . The system of claim 1 , wherein the one or more processors are configured to:
receive a set of opt-in messages associated with the set of users, wherein the set of first nodes are generated in response to the set of opt-in messages.

4 . The system of claim 1 , wherein the one or more processors are configured to:
receive a set of weights associated with the set of entities, wherein the weighted distances are calculated using the set of weights.

5 . The system of claim 1 , wherein the one or more processors are configured to:
identify at least one relevant entity, represented by a second node in the portion of the set of second nodes, as a nearest neighbor using the weighted distances, wherein the at least one communication is associated with the at least one relevant entity.

6 . The system of claim 1 , wherein the one or more processors, to generate the at least one communication, are configured to:
select the at least one communication, from a plurality of candidate communications, using the weighted distances.

7 . The system of claim 1 , wherein the event information is associated with a first time period, and the at least one communication is associated with a second time period subsequent to the first time period.

8 . A method of applying rules to improve nearest neighbor matching, comprising:
receiving, from a data source, a set of first data structures associated with a set of users; receiving, from the data source, a set of second data structures associated with a set of entities; receiving event information associated with the set of users and the set of entities; generating, by a machine learning system, an embedding space that represents the set of first data structures, the set of second data structures, and the event information; disregarding a portion of the set of second data structures, for a selected user in the set of users, using a condition in order to generate a subset of the set of second data structures; identifying at least one relevant entity, from the subset of the set of second data structures, using the embedding space; generating, by the machine learning system, at least one communication associated with the at least one relevant entity; and outputting the at least one communication.

9 . The method of claim 8 , further comprising:
receiving, from an administrator device, an indication of the condition.

10 . The method of claim 8 , further comprising:
receiving, from an administrator device, a data structure encoding the at least one communication.

11 . The method of claim 8 , further comprising:
receiving, from an administrator device, a data structure encoding a plurality of candidate communications, wherein the at least one communication is generated from the plurality of candidate communications.

12 . The method of claim 8 , further comprising:
transmitting, to an administrator device, an indication that the at least one communication was output.

13 . The method of claim 8 , further comprising:
transmitting, to an administrator device, an indication of an amount associated with output of the at least one communication.

14 . The method of claim 8 , wherein outputting the at least one communication comprises:
transmitting the at least one communication to a user device associated with the selected user.

15 . A non-transitory computer-readable medium storing a set of instructions for configuring rules to improve nearest neighbor matching, the set of instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the device to:
transmit, to a remote system, a registration message that authorizes a remote system to access event information;
transmit, to the remote system, an indication of a condition;
transmit, to the remote system, a data structure encoding at least one communication; and
receive, from the remote system, a confirmation that the at least one communication was sent to a selected user, based on an embedding space, that satisfies the condition.

16 . The non-transitory computer-readable medium of claim 15 , wherein the registration message includes a set of credentials associated with the event information.

17 . The non-transitory computer-readable medium of claim 15 , wherein the indication of the condition includes a selection of the condition from a plurality of candidate conditions.

18 . The non-transitory computer-readable medium of claim 15 , wherein the confirmation indicates the selected user.

19 . The non-transitory computer-readable medium of claim 15 , wherein the one or more instructions, when executed by the one or more processors, cause the device to:
receive, from the remote system, an indication of an amount associated with transmission of the at least one communication.

20 . The non-transitory computer-readable medium of claim 15 , wherein the one or more instructions, when executed by the one or more processors, further cause the device to:
transmit, to the remote system, an indication of a geographic area, wherein the confirmation is received based on the selected user being associated with the geographic area.

Assignees

• Capital One Services Llc

Inventors

• Reinert Owen
• Goodsitt Jeremy
• Turner Taylor
• Au-Yeung Justin

Classifications

• G06N20/00

  Machine learning · CPC title

• G06N5/022 Primary

  Knowledge engineering; Knowledge acquisition · CPC title