Granted Patent B2

US 12,524,496 · App. 17/648,753 · Granted Jan 13, 2026

Inventors: Sundarrajan Raman (Chennai, IN); Hemamalini Rajagopalan (Chennai, IN)

Assignee: Bank of America Corporation

G06F18/2148G06F15/16G06F21/6218G06N10/40G06F2221/2141

Abstract

Claims (100)

1 . An access control device, comprising:

a memory operable to store a plurality of machine learning models, wherein each machine learning model is configured to:

receive user information identifying a user and a data identifier identifying a data element; and

output an access response indicating whether the user is approved for accessing the data element; and

a processor operably coupled to the memory, configured to:

receive a first training data comprising a first set of user information for a first plurality of users and a first set of data identifiers for a first plurality of data elements;

determine a first data size for the first training data based at least in part on a number of entries in the first training data;

compare the first data size to a predetermined data size threshold value;

determine the first data size is greater than the predetermined data size threshold value;

send the first training data to a quantum computing device in response to determining that the first data size is greater than the predetermined data size threshold value, wherein:

the quantum computing device is configured to process data using quantum bits; and

the quantum computing device is configured to train a first machine learning model using the first training data;

receive a first set of machine learning model parameters comprising a first set of weight coefficients from the quantum computing device in response to training the first machine learning model, wherein the first set of machine learning model parameters correspond with a global minima value for the first training data;

configure a second machine learning model from among the plurality of machine learning models using the first set of machine learning model parameters;

receive an access control request comprising a plurality of entries, wherein each entry identifies:

data identifiers for a plurality of data elements; and

user information comprising user identifiers for a plurality of users, wherein each user identifier is linked with one or more data identifiers for data elements a corresponding user is requesting permission to access;

determine if a second data size of the access control request is greater than the predetermined data size threshold value, wherein the second data size is determined based on a number of entries in the access control request; and

in response to determining that the second data size of the access control request is greater than the predetermined data size threshold value, utilize the second machine learning model to determine if each of the user identifiers is approved to access a corresponding data element linked with the one or more data identifiers, wherein the second machine learning model is configured with the first set of machine learning model parameters that correspond with the global minima value.

2 . The device of claim 1 , wherein the processor is further configured to:

receive a second training data comprising a second set of user information for a second plurality of users and a second set of data identifiers for a second plurality of data elements;

determine a third data size for the second training data based at least in part on a number of entries in the second training data;

compare the third data size to the predetermined data size threshold value;

determine the third data size is less than or equal to the predetermined data size threshold value;

train a third machine learning model locally using the second training data in response to determining that the first data size is less than or equal to the predetermined data size threshold value, wherein training the third machine learning model locally generates a second set of machine learning model parameters comprising a second set of weight coefficients that correspond with a local minima value for the second training data; and

configure a fourth machine learning model from among the plurality of machine learning models using the second set of machine learning model parameters.

3 . The device of claim 1 , wherein the processor is further configured to:

select a machine learning model from among the plurality of machine learning models based on the number of entries in the access control request, wherein:

the selected machine learning model is configured to use machine learning model parameters for a local minima when the number of entries in the access control request is less than or equal to the predetermined data size threshold value;

input the user information and the data identifiers into the selected machine learning model; and

obtain an access response from the selected machine learning model in response to inputting the user information and the data identifiers into the selected machine learning model, wherein the access response indicates whether each user identifier is approved for accessing a data element.

4 . The device of claim 3 , wherein the access response comprises machine-executable instructions for decrypting an encrypted data element within a digital document.

5 . The device of claim 3 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to a data element within the digital document.

6 . The device of claim 3 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to the digital document.

7 . The device of claim 1 , wherein the access response comprises machine-executable instructions for modifying permission settings on a user device for the user.

8 . A data access control method, comprising:

receiving a first training data comprising a first set of user information for a first plurality of users and a first set of data identifiers for a first plurality of data elements;

determining a first data size for the first training data based at least in part on a number of entries in the first training data;

comparing the first data size to a predetermined data size threshold value;

determining the first data size is greater than the predetermined data size threshold value;

sending the first training data to a quantum computing device in response to determining that the first data size is greater than the predetermined data size threshold value, wherein:

the quantum computing device is configured to process data using quantum bits; and

the quantum computing device is configured to train a first machine learning model using the first training data;

receiving a first set of machine learning model parameters comprising a first set of weight coefficients from the quantum computing device in response to training the first machine learning model, wherein the first set of machine learning model parameters correspond with a global minima value for the first training data; and

configuring a second machine learning model from among a plurality of machine learning models using the first set of machine learning model parameters, wherein the second machine learning model is configured to:

receive user information identifying a user and a data identifier identifying a data element; and

output an access response indicating whether the user is approved for accessing the data element;

receiving an access control request comprising a plurality of entries, wherein each entry identifies:

data identifiers for a plurality of data elements; and

user information comprising user identifiers for a plurality of users, wherein each user identifier is linked with one or more data identifiers for data elements a corresponding user is requesting permission to access;

determining if a second data size of the access control request is greater than the predetermined data size threshold value, wherein the second data size is determined based on a number of entries in the access control request; and

in response to determining that the second data size of the access control request is greater than the predetermined data size threshold value, utilizing the second machine learning model to determine if each of the user identifiers is approved to access a corresponding data element linked with the one or more data identifiers, wherein the second machine learning model is configured with the first set of machine learning model parameters that correspond with the global minima value.

9 . The method of claim 8 , further comprising:

receiving a second training data comprising a second set of user information for a second plurality of users and a second set of data identifiers for a second plurality of data elements;

determining a third data size for the second training data based at least in part on a number of entries in the second training data;

comparing the third data size to the predetermined data size threshold value;

determining the third data size is less than or equal to the predetermined data size threshold value;

training a third machine learning model locally using the second training data in response to determining that the first data size is less than or equal to the predetermined data size threshold value, wherein training the third machine learning model locally generates a second set of machine learning model parameters comprising a second set of weight coefficients that correspond with a local minima value for the second training data; and

configuring a fourth machine learning model from among the plurality of machine learning models using the second set of machine learning model parameters.

10 . The method of claim 8 , further comprising:

selecting a machine learning model from among the plurality of machine learning models based on the number of entries in the access control request, wherein:

the selected machine learning model is configured to use machine learning model parameters for a local minima when the number of entries in the access control request is less than or equal to the predetermined data size threshold value;

inputting the user information and the data identifiers into the selected machine learning model; and

obtaining an access response from the selected machine learning model in response to inputting the user information and the data identifiers into the selected machine learning model, wherein the access response indicates whether each user identifier is approved for accessing a data element.

11 . The method of claim 10 , wherein the access response comprises machine-executable instructions for decrypting an encrypted data element within a digital document.

12 . The method of claim 10 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to a data element within the digital document.

13 . The method of claim 10 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to the digital document.

14 . The method of claim 10 , wherein the access response comprises machine-executable instructions for modifying permission settings on a user device for the user.

15 . A non-transitory computer-readable medium storing instructions that when executed by a processor cause the processor to:

receive a first training data comprising a first set of user information for a first plurality of users and a first set of data identifiers for a first plurality of data elements;

determine a first data size for the first training data based at least in part on a number of entries in the first training data;

compare the first data size to a predetermined data size threshold value;

determine the first data size is greater than the predetermined data size threshold value;

send the first training data to a quantum computing device in response to determining that the first data size is greater than the predetermined data size threshold value, wherein:

the quantum computing device is configured to process data using quantum bits; and

the quantum computing device is configured to train a first machine learning model using the first training data;

receive a first set of machine learning model parameters comprising a first set of weight coefficients from the quantum computing device in response to training the first machine learning model, wherein the first set of machine learning model parameters correspond with a global minima value for the first training data; and

configure a second machine learning model from among a plurality of machine learning models using the first set of machine learning model parameters, wherein the second machine learning model is configured to:

receive user information identifying a user and a data identifier identifying a data element; and

output an access response indicating whether the user is approved for accessing the data element;

receive an access control request comprising a plurality of entries, wherein each entry identifies:

data identifiers for a plurality of data elements; and

user information comprising user identifiers for a plurality of users, wherein each user identifier is linked with one or more data identifiers for data elements a corresponding user is requesting permission to access;

determine if a second data size of the access control request is greater than the predetermined data size threshold value, wherein the second data size is determined based on a number of entries in the access control request; and

in response to determining that the second data size of the access control request is greater than the predetermined data size threshold value, utilize the second machine learning model to determine if each of the user identifiers is approved to access a corresponding data element linked with the one or more data identifiers, wherein the second machine learning model is configured with the first set of machine learning model parameters that correspond with the global minima value.

16 . The non-transitory computer-readable medium of claim 15 , wherein the instructions further cause the processor to:

receive a second training data comprising a second set of user information for a second plurality of users and a second set of data identifiers for a second plurality of data elements;

determine a third data size for the second training data based at least in part on a number of entries in the second training data;

compare the third data size to the predetermined data size threshold value;

determine the third data size is less than or equal to the predetermined data size threshold value;

train a third machine learning model locally using the second training data in response to determining that the first data size is less than or equal to the predetermined data size threshold value, wherein training the third machine learning model locally generates a second set of machine learning model parameters comprising a second set of weight coefficients that correspond with a local minima value for the second training data; and

configure a fourth machine learning model from among the plurality of machine learning models using the second set of machine learning model parameters.

17 . The non-transitory computer-readable medium of claim 15 , wherein the instructions further cause the processor to:

select a machine learning model from among the plurality of machine learning models based on the number of entries in the access control request, wherein:

the selected machine learning model is configured to use machine learning model parameters for a local minima when the number of entries in the access control request is less than or equal to the predetermined data size threshold value;

input the user information and the data identifiers into the selected machine learning model; and

obtain an access response from the selected machine learning model in response to inputting the user information and the data identifiers into the selected machine learning model, wherein the access response indicates whether each user identifier is approved for accessing a data element.

18 . The non-transitory computer-readable medium of claim 17 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to a data element within the digital document.

19 . The non-transitory computer-readable medium of claim 17 , wherein the access response comprises machine-executable instructions for modifying a digital document to provide access to the digital document.

20 . The non-transitory computer-readable medium of claim 17 , wherein the access response comprises machine-executable instructions for modifying permission settings on a user device for the user.

Assignments (1)

Continuity (1)

References Cited (84)