Granted Patent B1

US 12,541,560 · App. 19/185,901 · Granted Feb 3, 2026

Inventors: Blake Browder (Dallas, TX); Joy Figarsky (Little Rock, AR)

Assignee: BH Operations, LLC

G06F16/906G06F16/9027G06F16/951G06N3/09G06N3/094G06N7/01

Abstract

Claims (52)

1 . An apparatus for generative interpolation, the apparatus comprising:

at least one processor; and

a memory communicatively connected to the at least one processor, wherein the memory contains instructions configuring the at least one processor to:

receive a plurality of attributes;

classify the plurality of attributes into one or more hierarchical groups;

generate detection training data, wherein the detection training data comprises correlations between classified attributes and exemplary missing attributes;

generate a detection machine-learning model using the detection training data;

update the detection training data using a feedback loop as a function of at least an output of the detection machine-learning model;

fine-tune the detection machine-learning model using the updated detection training data;

detect at least one missing attribute in the one or more hierarchical groups using the fine-tuned detection machine-learning model;

retrieve at least one crawled attribute as a function of the at least one missing attribute, wherein retrieving the at least one crawled attribute comprises:

dynamically loading contents for retrieving the at least one crawled attribute from web sources; and

updating the one or more hierarchical groups as a function of the at least one crawled attribute; and

generate a hierarchical data structure as a function of the one or more updated hierarchical groups.

2 . The apparatus of claim 1 , wherein the plurality of attributes are associated with inventory of a facility.

3 . The apparatus of claim 2 , wherein the plurality of attributes are associated with financial statements of a hospital.

4 . The apparatus of claim 1 , wherein receiving the plurality of attributes comprises:

receiving input data from one or more data sources; and

extracting the plurality of attributes from the input data.

5 . The apparatus of claim 1 , wherein classifying the plurality of attributes into the one or more hierarchical groups comprises classifying the plurality of attributes using a trained group classifier.

6 . The apparatus of claim 5 , wherein classifying the plurality of attributes comprises training the group classifier, comprising:

generating classification training data, wherein the classification training data comprises exemplary attributes correlated to exemplary hierarchical groups; and

training the group classifier using the classification training data.

7 . The apparatus of claim 1 , wherein the at least one missing attribute is detected when an attribute of the plurality of attributes required to be in the one or more hierarchical group is absent.

8 . The apparatus of claim 1 , wherein dynamically loading the contents for retrieving the at least one crawled attribute from the web sources comprises dynamically loading the contents for retrieving the at least one crawled attribute from the web sources until the at least one crawled attribute is found to fill the at least one missing attribute in the one or more hierarchical groups.

9 . The apparatus of claim 1 , wherein the processor is further configured to present the hierarchical data structure through a graphical user interface.

10 . The apparatus of claim 9 , wherein the hierarchical data structure comprises a visual indicator for at least one missing attribute.

11 . A method for generative interpolation, the method comprising:

receiving, using at least a processor, a plurality of attributes;

classifying, using the at least a processor, the plurality of attributes into one or more hierarchical groups;

generating, using the at least a processor, detection training data wherein the detection training data comprises correlations between classified attributes and exemplary missing attributes;

generating, using the at least a processor, a detection machine-learning model using the detection training data;

updating, using the at least a processor, the detection training data using a feedback loop as a function of at least an output of the detection machine-learning model;

fine-tuning, using the at least a processor, the detection machine-learning model using the updated detection training data;

detecting, using the at least a processor, at least one missing attribute in the one or more hierarchical groups using the fine-tuned detection machine-learning model;

retrieving, using the at least a processor, at least one crawled attribute as a function of the at least one missing attribute, wherein retrieving the at least one crawled attribute comprises:

dynamically loading contents for retrieving the at least one crawled attribute from web sources; and

updating the one or more hierarchical groups as a function of the at least one crawled attribute; and

generating, using the at least a processor, a hierarchical data structure as a function of the one or more updated hierarchical groups.

12 . The method of claim 11 , wherein the plurality of attributes are associated with inventory of a facility.

13 . The method of claim 12 , wherein the plurality of attributes are associated with financial statements of a hospital.

14 . The method of claim 11 , wherein receiving, using the at least a processor, the plurality of attributes comprises:

receiving input data from one or more data sources; and

extracting the plurality of attributes from the input data.

15 . The method of claim 11 , wherein classifying, using the at least a processor, the plurality of attributes into the one or more hierarchical groups comprises classifying the plurality of attributes using a trained group classifier.

16 . The method of claim 15 , wherein classifying, using the at least a processor, the plurality of attributes comprises training the group classifier comprising:

generating classification training data, wherein the classification training data comprises exemplary attributes correlated to exemplary hierarchical groups; and

training the group classifier using the classification training data.

17 . The method of claim 11 , wherein the at least one missing attribute is detected when an attribute of the plurality of attributes required to be in the one or more hierarchical group is absent.

18 . The method of claim 11 , wherein dynamically loading the contents for retrieving the at least one crawled attribute from the web sources comprises dynamically loading the contents for retrieving the at least one crawled attribute from the web sources until the at least one crawled attribute is found to fill the at least one missing attribute in the one or more hierarchical groups.

19 . The method of claim 11 , the method further comprising, presenting, using the at least a processor, the hierarchical data structure through a graphical user interface.

20 . The method of claim 19 , wherein the hierarchical data structure comprises a visual indicator for at least one missing attribute.

Assignments (3)

Continuity (1)

References Cited (13)