Granted Patent B2

US 12,567,039 · App. 16/380,363 · Granted Mar 3, 2026

Inventors: Leandro Eidelwein (Porto Alegre, BR); Bruna Gouveia (Porto Alegre, BR); Rafael Gomes (Porto Alegre, BR); Roberto Dias (Sao Paulo, BR); Andre Mendes (Porto Alegre, BR); Eduardo Hoefel (Porto Alegre, BR)

Assignee: ADP, Inc.

G06Q10/1053G06N20/00

Abstract

Claims (62)

1 . A computer-implemented method, comprising:

receiving, by one or more processors coupled with memory, current data and historic data associated with a unique data object, wherein the current data and the historic data are in an unstructured data format;

extracting, by the one or more processors, data values from the current data and the historic data;

generating, by the one or more processors, confirmed data values by performing disambiguation processes on the data values and by selecting semantic meanings of text of the data values from a plurality of possible meanings of word content as a first function of context of the data values;

mapping, by the one or more processors, the confirmed data values to data elements of a first metadata structure of a plurality of metadata structures, based on the confirmed data values matching one or more of the data elements of the first metadata structure, wherein each of the plurality of metadata structures are in a unified data format, and wherein the data elements of the first metadata structure include a title;

executing, by the one or more processors, a machine learning process by inputting the first metadata structure to the machine learning process, wherein the machine learning process compares the first metadata structure to the plurality of metadata structures to identify a subset of metadata structures of the plurality of metadata structures based on i) the data elements of each of the subset of metadata structures including values greater than the data elements of a remainder of the plurality of metadata structures, and ii) the data elements of each of the subset of metadata structures being similar to the data elements of the first metadata structure, and wherein each of the subset of metadata structures include a unique title different from the title of the first metadata structure;

determining, by the one or more processors, a viability value for each of the subset of metadata structures as a second function of strength of match the data elements of the first metadata structure to the data elements of each of the subset of metadata structures, wherein the greater the viability value, the greater a quantity of the data elements are shared by the first metadata structure and a respective metadata structure of the subset of metadata structures;

generating, by the one or more processors, a digital twin of the first metadata structure using principal component analysis (PCA) to perform a dimensionality reduction of the data elements of the first metadata structure, clustering the data elements of the first metadata structure with the data elements of the subset of metadata structures, wherein the PCA transforms the data elements of the first metadata structure into a set of linearly uncorrelated principal components;

executing, by the one or more processors, the digital twin of the first metadata structure in a simulated scenario of each of the subset of metadata structures to determine a progression value for each of the subset of metadata structures at an end of a time period;

determining, by the one or more processors, composite values for the subset of metadata structures based on the viability value and the progression value;

identifying, by the one or more processors, a first group including metadata structures of the subset of metadata structures including the composite values in a top third of all the composite values, a second group including the metadata structures of the subset of metadata structures including the composite values in a middle third of all the composite values, and a third group including the metadata structures of the subset of metadata structures including the composite values in a bottom third of all the composite values;

providing, by the one or more processors, the unique title of each of the metadata structures of the first group, the second group, and the third group for display using a graphical user interface, wherein the unique title of each of the metadata structures of the first group are displayed with a first font and a first color to indicate that the metadata structures of the first group are recommended for the unique data object, and wherein the unique title of each of the metadata structures of the second group and the third group are displayed in a font different from the first font and a color different from the first color to indicate that the metadata structures of the second group and the third group are not as recommended as the metadata structures of the first group;

receiving, by the one or more processors, a selection via the graphical user interface of a metadata structure of the first group having the unique title displayed with the first font and the first color; and

automatically navigating, by the one or more processors, in response to the selection via the graphical user interface associated with the first color, to a browser to provide access to an application associated with the selection.

2 . A system, comprising:

a processor;

a computer readable memory in circuit communication with the processor;

a computer readable storage medium in circuit communication with the processor;

wherein the processor executes program instructions stored on the computer readable storage medium via the computer readable memory and thereby:

receives current data and historic data associated with a unique data object, wherein the current data and the historic data are in an unstructured data format;

extracts data values from the current data and the historic data;

generates confirmed data values by performing disambiguation processes on the data values and by selecting semantic meanings of text of the data values from a plurality of possible meanings of word content as a first function of context of the data values;

maps, the confirmed data values to data elements of a first metadata structure of a plurality of metadata structures, based on the confirmed data values matching one or more of the data elements of the first metadata structure, wherein each of the plurality of metadata structures are in a unified data format, and wherein the data elements of the first metadata structure include a title;

executes a machine learning process by inputting the first metadata structure to the machine learning process, wherein the machine learning process compares the first metadata structure to the plurality of metadata structures to identify a subset of metadata structures of the plurality of metadata structures based on i) the data elements of each of the subset of metadata structures including values greater than the data elements of a remainder of the plurality of metadata structures, and ii) the data elements of each of the subset of metadata structures being similar to the data elements of the first metadata structure, and wherein each of the subset of metadata structures include a unique title different from the title of the first metadata structure;

determines a viability value for each of the subset of metadata structures as a second function of strength of match the data elements of the first metadata structure to the data elements of each of the subset of metadata structures, wherein the greater the viability value, the greater a quantity of the data elements are shared by the first metadata structure and a respective metadata structure of the subset of metadata structures;

generates a digital twin of the first metadata structure using principal component analysis (PCA) to perform a dimensionality reduction of the data elements of the first metadata structure, clustering the data elements of the first metadata structure with the data elements of the subset of metadata structures, wherein the PCA transforms the data elements of the first metadata structure into a set of linearly uncorrelated principal components;

executes the digital twin of the first metadata structure in a simulated scenario of each of the subset of metadata structures to determine a progression value for each of the subset of metadata structures at an end of a time period;

determines composite values for the subset of metadata structures based on the viability value and the progression value;

identifies a first group including metadata structures of the subset of metadata structures including the composite values in a top third of all the composite values, a second group including the metadata structures of the subset of metadata structures including the composite values in a middle third of all the composite values, and a third group including the metadata structures of the subset of metadata structures including the composite values in a bottom third of all the composite values;

provides the unique title of each of the metadata structures of the first group, the second group, and the third group for display using a graphical user interface, wherein the unique title of each of the metadata structures of the first group are displayed with a first font and a first color to indicate that the metadata structures of the first group are recommended for the unique data object, and wherein the unique title of each of the metadata structures of the second group and the third group are displayed in a font different from the first font and a color different from the first color to indicate that the metadata structures of the second group and the third group are not as recommended as the metadata structures of the first group;

receives a selection via the graphical user interface of a metadata structure of the first group having the unique title displayed with the first font and the first color; and

automatically navigates, in response to the selection via the graphical user interface associated with the first color, to a browser to provide access to an application associated with the selection.

3 . A computer program product, comprising:

a computer readable storage medium having computer readable program code embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the computer readable program code comprising instructions for execution by a processor that cause the processor to:

receive current data and historic data associated with a unique data object, wherein the current data and the historic data are in an unstructured data format;

extract data values from the current data and the historic data;

generate confirmed data values by performing disambiguation processes on the data values and by selecting semantic meanings of text of the data values from a plurality of possible meanings of word content as a first function of context of the data values;

map the confirmed data values to data elements of a first metadata structure of a plurality of metadata structures, based on the confirmed data values matching one or more of the data elements of the first metadata structure, wherein each of the plurality of metadata structures are in a unified data format, and wherein the data elements of the first metadata structure include a title;

execute a machine learning process by inputting the first metadata structure to the machine learning process, wherein the machine learning process compares the first metadata structure to the plurality of metadata structures to identify a subset of metadata structures of the plurality of metadata structures based on i) the data elements of each of the subset of metadata structures including values greater than the data elements of a remainder of the plurality of metadata structures, and ii) the data elements of each of the subset of metadata structures being similar to the data elements of the first metadata structure, and wherein each of the subset of metadata structures include a unique title different from the title of the first metadata structure;

determine a viability value for each of the subset of metadata structures as a second function of strength of match the data elements of the first metadata structure to the data elements of each of the subset of metadata structures, wherein the greater the viability value, the greater a quantity of the data elements are shared by the first metadata structure and a respective metadata structure of the subset of metadata structures;

generate a digital twin of the first metadata structure using principal component analysis (PCA) to perform a dimensionality reduction of the data elements of the first metadata structure, clustering the data elements of the first metadata structure with the data elements of the subset of metadata structures, wherein the PCA transforms the data elements of the first metadata structure into a set of linearly uncorrelated principal components;

execute the digital twin of the first metadata structure in a simulated scenario of each of the subset of metadata structures to determine a progression value for each of the subset of metadata structures at an end of a time period;

determine composite values for the subset of metadata structures based on the viability value and the progression value;

identify a first group including metadata structures of the subset of metadata structures including the composite values in a top third of all the composite values, a second group including the metadata structures of the subset of metadata structures including the composite values in a middle third of all the composite values, and a third group including the metadata structures of the subset of metadata structures including the composite values in a bottom third of all the composite values;

provide the unique title of each of the metadata structures of the first group, the second group, and the third group for display using a graphical user interface, wherein the unique title of each of the metadata structures of the first group are displayed with a first font and a first color to indicate that the metadata structures of the first group are recommended for the unique data object, and wherein the unique title of each of the metadata structures of the second group and the third group are displayed in a font different from the first font and a color different from the first color to indicate that the metadata structures of the second group and the third group are not as recommended as the metadata structures of the first group;

receive a selection via the graphical user interface of a metadata structure of the first group having the unique title displayed with the first font and the first color; and

automatically navigate, in response to the selection via the graphical user interface associated with the first color, to a browser to provide access to an application associated with the selection.

4 . The method of claim 1 , comprising:

receiving, by the one or more processors, an indication of a selection of a first unique title of a second metadata structure of at least one of the first group, the second group, and the third group;

identifying, by the one or more processors, responsive to the selection, a plurality of data elements that are included in the data elements of the second metadata structure and missing from the data elements of the first metadata structure;

determining, by the one or more processors, a plurality of viability values for the second metadata structure as a third function of changes to the plurality of data elements; and

providing, by the one or more processors, a list comprising the plurality of data elements for display using the graphical user interface, wherein the plurality of data elements of the list are ranked based on the plurality of viability values.

5 . The system of claim 2 , comprising:

receives an indication of a selection of a first unique title of a second metadata structure of at least one of the first group, the second group, and the third group;

identifies, responsive to the selection, a plurality of data elements that are included in the data elements of the second metadata structure and missing from the data elements of the first metadata structure;

determines a plurality of viability values for the second metadata structure as a third function of changes to the plurality of data elements; and

provides a list comprising the plurality of data elements for display using the graphical user interface, wherein the plurality of data elements of the list are ranked based on the plurality of viability values.

6 . The computer program product of claim 3 , wherein the computer readable program code instructions for execution by the processor further cause the processor to:

receive an indication of a selection of a first unique title of a second metadata structure of at least one of the first group, the second group, and the third group;

identify, responsive to the selection, a plurality of data elements that are included in the data elements of the second metadata structure and missing from the data elements of the first metadata structure;

determine a plurality of viability values for the second metadata structure as a third function of changes to the plurality of data elements; and

provide a list comprising the plurality of data elements for display using the graphical user interface, wherein the plurality of data elements of the list are ranked based on the plurality of viability values.

Assignments (2)

Continuity (1)

References Cited (29)