Dynamically generating activity prompts to build and refine machine learning authentication models

Patent metadata

Field	Value
Publication number	US-11048793-B2
Application number	US-201816210092-A
Country	US
Kind code	B2
Filing date	Dec 5, 2018
Priority date	Dec 5, 2018
Publication date	Jun 29, 2021
Grant date	Jun 29, 2021

Abstract

Aspects of the disclosure relate to dynamically generating activity prompts to build and refine machine learning authentication models. A computing platform may process a first set of login events associated with a first user account and may build a first user-specific authentication model for the first user account. Then, the computing platform may process a second set of login events associated with a second user account and may build a second user-specific authentication model for the second user account. The computing platform also may build a population-level authentication model for a plurality of user accounts. Thereafter, the computing platform may identify one or more activity parameters associated with at least one authentication model for refinement. Subsequently, the computing platform may generate and send one or more activity prompts to one or more client computing devices to request at least one user response.

First claim

What is claimed is:

1. A computing platform, comprising:
at least one processor;
a communication interface communicatively coupled to the at least one processor; and
memory storing computer-readable instructions that, when executed by the at least one processor, cause the computing platform to:
process a first set of login events associated with a first user account;
build a first user-specific authentication model for the first user account based on first activity data captured during the first set of login events associated with the first user account;
process a second set of login events associated with a second user account;
build a second user-specific authentication model for the second user account based on second activity data captured during the second set of login events associated with the second user account;
build a population-level authentication model for a plurality of user accounts based on the first activity data captured during the first set of login events associated with the first user account and the second activity data captured during the second set of login events associated with the second user account;
identify one or more activity parameters associated with at least one authentication model for refinement;
generate one or more activity prompts based on the one or more activity parameters associated with the at least one authentication model identified for refinement; and
send, via the communication interface, to at least one client computing device, at least one activity prompt of the one or more activity prompts, wherein sending the at least one activity prompt of the one or more activity prompts causes the at least one client computing device to request at least one user response corresponding to the at least one activity prompt of the one or more activity prompts.

2. The computing platform of claim 1 , wherein processing the first set of login events associated with the first user account comprises:
receiving, via the communication interface, from an account portal computing platform, a first authentication request associated with a first client computing device requesting access to one or more secured information resources associated with the first user account;
validating one or more authentication credentials associated with the first authentication request; and
based on validating the one or more authentication credentials associated with the first authentication request, generating one or more authentication commands directing the account portal computing platform to provide the first client computing device with access to the one or more secured information resources associated with the first user account.

3. The computing platform of claim 1 , wherein building the first user-specific authentication model for the first user account comprises:
receiving parameter-based activity data associated with at least two login events involving the first user account;
calculating one or more profile values for one or more authentication parameters based on the parameter-based activity data associated with the at least two login events involving the first user account; and
storing the one or more profile values calculated for the one or more authentication parameters in a dataset defining the first user-specific authentication model.

4. The computing platform of claim 1 , wherein the first activity data captured during the first set of login events associated with the first user account comprises information identifying one or more device identifiers, one or more network addresses, one or more internet service providers, one or more device versions, one or more device operating systems, one or more device operating system versions, one or more device locations, one or more requested channels, one or more requested pages, one or more sequences of page requests, or one or more timings between page requests.

5. The computing platform of claim 1 , wherein processing the second set of login events associated with the second user account comprises:
receiving, via the communication interface, from an account portal computing platform, a second authentication request associated with a second client computing device requesting access to one or more secured information resources associated with the second user account;
validating one or more authentication credentials associated with the second authentication request; and
based on validating the one or more authentication credentials associated with the second authentication request, generating one or more authentication commands directing the account portal computing platform to provide the second client computing device with access to the one or more secured information resources associated with the second user account.

6. The computing platform of claim 1 , wherein building the second user-specific authentication model for the second user account comprises:
receiving parameter-based activity data associated with at least two login events involving the second user account;
calculating one or more profile values for one or more authentication parameters based on the parameter-based activity data associated with the at least two login events involving the second user account; and
storing the one or more profile values calculated for the one or more authentication parameters in a dataset defining the second user-specific authentication model.

7. The computing platform of claim 1 , wherein the second activity data captured during the second set of login events associated with the second user account comprises information identifying one or more device identifiers, one or more network addresses, one or more internet service providers, one or more device versions, one or more device operating systems, one or more device operating system versions, one or more device locations, one or more requested channels, one or more requested pages, one or more sequences of page requests, or one or more timings between page requests.

8. The computing platform of claim 1 , wherein building the population-level authentication model for the plurality of user accounts comprises:
receiving parameter-based activity data associated with at least two login events involving at least two different user accounts;
calculating one or more population-target values for one or more authentication parameters based on the parameter-based activity data associated with the at least two login events involving the at least two different user accounts; and
storing the one or more population-target values calculated for the one or more authentication parameters in a dataset defining the population-level authentication model.

9. The computing platform of claim 1 , wherein identifying the one or more activity parameters associated with the at least one authentication model for refinement comprises identifying a first activity parameter for refinement in response to determining that an amount of measured data associated with the first activity parameter is below a predetermined threshold corresponding to the first activity parameter.

10. The computing platform of claim 1 , wherein identifying the one or more activity parameters associated with the at least one authentication model for refinement comprises identifying a second activity parameter for refinement in response to determining that measured data associated with the second activity parameter has expired.

11. The computing platform of claim 1 , wherein identifying the one or more activity parameters associated with the at least one authentication model for refinement comprises determining that the first user-specific authentication model for the first user account or the second user-specific au

Assignees

• Bank Of America

Inventors

• Toth Michael E
• Shah Hitesh

Classifications

• G06F21/31

  User authentication · CPC title

• G06F21/45 Primary

  Structures or tools for the administration of authentication · CPC title

• G06N20/00 Primary

  Machine learning · CPC title