Granted Patent B1

US 12,332,921 · App. 18/381,999 · Granted Jun 17, 2025

Inventors: Heather Marie Gavlak (Delaware, OH); Richard Gregory Barker (Gibsonia, PA); Evan Michael Sorger (Berea, OH); Kevin William Black (Pittsburgh, PA)

Assignee: THE PNC FINANCIAL SERVICES GROUP, INC.

G06F16/288G06F16/285G06F16/9024G06N20/00

Abstract

Claims (35)

1. A computer-system for maintaining a relational data model for storage in a database, comprising:

a storage device that stores instructions; and

at least one processor that executes the instructions to:

generate, using a machine learning model, a data analysis algorithm trained to predict relationships in a data domain according to an architectural framework, wherein the architectural framework includes four levels: a business level, an application level, a data level and a technology level;

logically segregating, using the data analysis algorithm, each data domain based on the architectural framework, wherein the logical segregation includes organizing data within each data domain into one of the four levels of the architectural framework;

access a graph database that includes nodes and edges as a data model for representing data objects of the data domain according to the architectural framework, the nodes being configured to store first-class entities and the edges being configured to store relationships between the first-class entities;

store individual data objects of the data domain within the database according to relationships between the first-class entities in the data domain, the relationships containing architectural properties of the architectural framework, for an instance of querying, the querying including searching for applications within the data domain;

query the data objects using a modeling language for the database;

apply the data analysis algorithm to identify patterns in the graph database; and

draw the identified graph patterns on the graph database.

2. The system of claim 1 , wherein the nodes include a set of data in the graph database for the data representation model of the graph database.

3. The system of claim 1 , wherein the relational metamodel dataset requires one or more disparate data tables to express the graph databases' capabilities in a modeled instance.

4. The system of claim 1 , wherein the graph database classifies relationships for the data objects as the first-class entities.

5. The system of claim 1 , wherein the graph database includes the relationships between data points that contain properties for the first-class entities.

6. The system of claim 1 , wherein the graph database includes querying data containing the relationships.

7. The system of claim 1 , wherein the data analysis algorithm includes: a tangible similarity algorithm and a tangible centrality algorithm.

8. The system of claim 7 , wherein the tangible centrality algorithm includes determining unique ones of the nodes within the graph database.

9. The system of claim 7 , wherein the tangible similarity algorithm includes computing similar pairs of the nodes based on their properties or properties of neighboring ones of the nodes.

10. A computer-implemented method for maintaining a relational data model for storage in a database, the method comprising the following operations performed by at least one processor:

at least one processor that executes the instructions to:

generating, using a machine learning model, a data analysis algorithm trained to predict relationships in a data domain according to an architectural framework, wherein the architectural framework includes four levels: a business level, an application level, a data level and a technology level,

logically segregating, using the data analysis algorithm, each data domain based on the architectural framework, wherein the logical segregation includes organizing data within each data domain into one of the four levels of the architectural framework;

accessing, a graph database that includes nodes and edges as a data model for representing data objects of the data domain according to the architectural framework, the nodes being configured to store first-class entities and the edges being configured to store relationships between the first-class entities;

storing, individual data objects of the data domain within the database according to relationships between the first-class entities in the data domain, the relationships containing architectural properties of the architectural framework, for an instance of querying, the querying including searching for applications within the data domain;

querying, the data objects using a modeling language for the database;

applying, the data analysis algorithm to identify patterns in the graph database; and

drawing the identified graph patterns on the graph database.

11. The system of claim 10 , wherein the nodes include the set of data in the graph database for the data representation model of the graph database.

12. The system of claim 10 , wherein the relational metamodel dataset requires one or more disparate data tables to express the graph databases' capabilities in a modeled instance.

13. The system of claim 10 , wherein the graph database of the set of data involves classifying relationships as the first-class entities.

14. The system of claim 10 , wherein the graph database includes the relationships between data points that contain properties for the first-class entities.

15. The system of claim 10 , wherein the graph database includes querying data containing the relationships.

16. The system of claim 10 , wherein the data analysis algorithm includes: a tangible similarity algorithm and a tangible centrality algorithm.

17. The system of claim 16 , wherein the tangible centrality algorithm includes determining unique ones of the nodes within the graph database.

18. The system of claim 16 , wherein the tangible similarity algorithm includes computing similar pairs of the nodes based on their properties or properties of neighboring ones of the nodes.

Assignments (1)

Continuity (1)

References Cited (76)