Automated resolution of over and under-specification in a knowledge graph

Patent metadata

Field	Value
Publication number	US-11741379-B2
Application number	US-202217897923-A
Country	US
Kind code	B2
Filing date	Aug 29, 2022
Priority date	May 29, 2019
Publication date	Aug 29, 2023
Grant date	Aug 29, 2023

Abstract

Systems and methods for automated resolution of over-specification and under-specification in a knowledge graph are disclosed. In embodiments, a method includes: determining, by a computing device, that a size of an object cluster of a knowledge graph meets a threshold value indicating under-specification of a knowledge base of the knowledge graph; determining, by the computing device, sub-classes for objects of the knowledge graph; re-initializing, by the computing device, the knowledge graph based on the sub-classes to generate a refined knowledge graph, wherein the size of the object cluster is reduced in the refined knowledge graph; and generating, by the computing device, an output based on information determined from the refined knowledge graph.

First claim

What is claimed is:

1. A system comprising:
a processor, a computer readable memory and a computer readable storage medium associated with a computing device;
program instructions to determine over-specification of a knowledge graph generated for a knowledge base, based on one or more object clusters of the knowledge graph having less than a threshold amount of data points; and
program instructions to re-initialize the knowledge graph based on the determining the over-specification to generate an updated knowledge graph wherein entities of the one or more object clusters of the knowledge graph are subsumed into a single object cluster of the updated knowledge graph,
wherein the program instructions are stored on the computer readable storage medium for execution by the processor via the computer readable memory.

2. The system of claim 1 , wherein the determining over-specification comprises performing a clustering analysis.

3. The system of claim 2 , wherein the clustering analysis utilizes a Voronoi cells cluster initialization formula.

4. The system of claim 1 , further comprising program instructions to generate the knowledge graph from the knowledge base.

5. The system of claim 1 , wherein the knowledge graph comprises networks of entities, semantic types, properties, and relationships between the entities.

6. The system of claim 1 , further comprising program instructions to determine objects of the one or more object clusters.

7. The system of claim 1 , further comprising program instructions to generate an output based on the updated knowledge graph.

8. The system of claim 7 , wherein the output is a report, the system further comprising program instructions to send the report to a remote client device via a network connection.

9. The system of claim 1 , wherein the threshold amount of data points is a threshold amount of data points for a domain type.

10. A system comprising:
a processor, a computer readable memory and a computer readable storage medium associated with a computing device;
program instructions to determine whether a size of an object cluster of a knowledge graph meets a threshold value indicating under-specification of a knowledge base of the knowledge graph;
program instructions to determine sub-classes for objects of the knowledge graph; and
program instructions to re-initialize the knowledge graph based on the sub-classes to generate a refined knowledge graph, wherein the size of the object cluster is reduced in the refined knowledge graph,
wherein the program instructions are stored on the computer readable storage medium for execution by the processor via the computer readable memory.

11. The system of claim 10 , further comprising program instructions to generate an output based on information determined from the refined knowledge graph.

12. The system of claim 11 , wherein the output is a report, the system further comprising program instructions to send the report to a remote client device via a network connection.

13. The system of claim 10 , wherein the determining the under-specification of the knowledge base comprises:
determining a value by subtracting an actual split in the object cluster from a hypothetical split in the object cluster; and
determining the under-specification of the knowledge base based on the value being greater than or equal to zero and greater than or equal to the threshold value indicating under-specification.

14. The system of claim 10 , wherein the determining the sub-classes for the knowledge graph comprises determining parent-child relationships within a latent hierarchical structure of the knowledge base.

15. The system of claim 10 , wherein the determining the sub-classes for the knowledge graph comprises determining the sub-classes for the knowledge graph based on peer nodes.

16. The system of claim 10 , wherein the determining the sub-classes comprises performing pattern matching using natural language processing to identify objects having a calculated node similarity meeting a threshold value of node similarity.

17. The system of claim 10 , further comprising program instructions to determine over-specification of the knowledge base by performing a clustering analysis.

18. The system of claim 17 , wherein the clustering analysis utilizes a Voronoi cells cluster initialization formula.

19. The system of claim 17 , further comprising program instructions to re-initializing the refined knowledge graph based on the determining the over-specification to generate an updated knowledge graph wherein entities of over-specified object clusters of the refined knowledge graph are subsumed into a single cluster of the updated knowledge graph.

20. The system of claim 10 , wherein the determining the over-specification is based on a threshold value of data points for a domain type.

Assignees

• Kyndryl Inc

Inventors

• Trim Craig M
• Rudden Mary E
• Marzorati Mauro
• Fox Jeremy R

Classifications

• G06N5/022 Primary

  Knowledge engineering; Knowledge acquisition · CPC title

• G06F18/2115

  by evaluating different subsets according to an optimisation criterion, e.g. class separability, forward selection or backward elimination · CPC title

• G06F18/2185

  the supervisor being an automated module, e.g. intelligent oracle · CPC title

• G06F18/231

  Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram · CPC title

• G06N5/04

  Inference or reasoning models · CPC title