Granted Patent B2

US 12,645,694 · App. 18/791,269 · Granted Jun 2, 2026

Inventors: Ranadhir Ghosh (St. Johns, FL); John Platais (Menomonee Falls, WI); Anup Patel (Bangalore, IN)

Assignee: Fidelity Information Services, LLC

G06F16/258G06F16/2237

Abstract

Claims (78)

1 . A computer-implemented method for finding historically similar incidents in a system, the method comprising:

receiving a plurality of historical data objects corresponding to a plurality of previous incidents, each of the plurality of historical data objects indicating an occurrence of a previous incident and including previous incident multivariate information;

processing the plurality of historical data objects;

converting the plurality of processed historical data objects to a plurality of historical embeddings by utilizing a large language model;

storing the plurality of historical embeddings in an index, the index being a vector embedding database;

receiving a data object indicating an occurrence of a current incident associated with a configurable item, the data object including multivariate information, the multivariate information including multiple data types, the data types including integers, free field text, and categories/subcategories;

processing the data object;

converting the data object into a first embedding by utilizing the large language model;

identifying a set of historically similar incidents by applying a Euclidean distance formula to the first embedding and the plurality of historical embeddings;

determining a similarity score for each of the set of historically similar incidents based on the application of the Euclidean distance formula; and

outputting to a display, a chart with information regarding the set of historically similar incidents, wherein the information includes indications regarding severities of the set of historically similar incidents, and the similarity scores of the set of historically similar incidents.

2 . The method of claim 1 , wherein the multivariate information includes a short description, a business category, a business sub-category, a line of business, and an incident priority.

3 . The method of claim 1 , wherein processing the data object includes:

formatting the multivariate information into a first string; and

performing preprocessing on the first string.

4 . The method of claim 3 , wherein performing preprocessing on the first string includes applying one or more of a lower casing algorithm, a tokenization algorithm, a punctuation mark removal algorithm, a stop word removal algorithm, a stemming algorithm, or a lemmatization algorithm.

5 . The method of claim 1 , wherein the similarity score is determined by applying exponential decay function to an output of the Euclidean distance formula for the set of historically similar incidents.

6 . The method of claim 1 , further including:

applying a normalizing algorithm to the similarity scores;

comparing the similarity scores to a threshold value;

determining a list of historically similar incidents with similarity scores greater than the threshold value; and

outputting the list of historically similar incidents to a user via a graphical user interface (GUI).

7 . A system for finding historically similar incidents in a system, the system comprising:

a memory having processor-readable instructions stored therein; and

at least one processor configured to access the memory and execute the processor-readable instructions to perform operations including:

receiving a plurality of historical data objects corresponding to a plurality of previous incidents, each of the plurality of historical data objects indicating an occurrence of a previous incident and including previous incident multivariate information;

processing the plurality of historical data objects;

converting the plurality of processed historical data objects to a plurality of historical embeddings by utilizing a large language model;

storing the plurality of historical embeddings in an index, the index being a vector embedding database;

receiving a data object indicating an occurrence of a current incident associated with a configurable item, the data object including multivariate information, the multivariate information including multiple data types, the data types including integers, free field text, and categories/subcategories;

processing the data object;

converting the data object into a first embedding by utilizing the large language model;

identifying a set of historically similar incidents by applying a Euclidean distance formula to the first embedding and the plurality of historical embeddings; and

determining a similarity score for each of the set of historically similar incidents based on the application of the Euclidean distance formula; and

outputting a signal configured to display on a display a chart information regarding the set of historically similar incidents, wherein the information includes indications regarding severities of the set of historically similar incidents, and the similarity scores of the set of historically similar incidents.

8 . The system of claim 7 , wherein the multivariate information includes a short description, a business category, a business sub-category, a line of business, and an incident priority.

9 . The system of claim 7 , wherein processing the data object includes:

formatting the multivariate information into a first string; and

performing preprocessing on the first string.

10 . The system of claim 9 , wherein performing preprocessing on the first string includes applying one or more of a lower casing algorithm, a tokenization algorithm, a punctuation mark removal algorithm, a stop word removal algorithm, a stemming algorithm, or a lemmatization algorithm.

11 . The system of claim 7 , wherein the similarity score is determined by applying exponential decay function to an output of the Euclidean distance formula for the set of historically similar incidents.

12 . The system of claim 7 , further including:

applying a normalizing algorithm to the similarity scores;

comparing the similarity scores to a threshold value;

determining a list of historically similar incidents with similarity scores greater than the threshold value; and

outputting the list of historically similar incidents to a user via a graphical user interface (GUI).

13 . A non-transitory computer readable medium storing processor-readable instructions which, when executed by at least one processor, cause the at least one processor to perform operations including:

receiving a plurality of historical data objects corresponding to a plurality of previous incidents, each of the plurality of historical data objects indicating an occurrence of a previous incident and including previous incident multivariate information;

processing the plurality of historical data objects;

converting the plurality of processed historical data objects to a plurality of historical embeddings by utilizing a large language model;

storing the plurality of historical embeddings in an index, the index being a vector embedding database;

receiving a data object indicating an occurrence of a current incident associated with a configurable item, the data object including multivariate information, the multivariate information including multiple data types, the data types including integers, free field text, and categories/subcategories;

processing the data object;

converting the data object into a first embedding by utilizing the large language model;

identifying a set of historically similar incidents by applying a Euclidean distance formula to the first embedding and the plurality of historical embeddings;

determining a similarity score for each of the set of historically similar incidents based on the application of the Euclidean distance formula; and

outputting a signal configured to cause a display to display a chart including information regarding the set of historically similar incidents, wherein the information includes indications regarding severities of the set of historically similar incidents, and the similarity scores of the set of historically similar incidents.

14 . The non-transitory computer readable medium of claim 13 , wherein the multivariate information includes a short description, a business category, a business sub-category, a line of business, and an incident priority.

15 . The non-transitory computer readable medium of claim 13 , wherein processing the data object includes:

formatting the multivariate information into a first string; and

performing preprocessing on the first string.

16 . The non-transitory computer readable medium of claim 15 , wherein performing preprocessing on the first string includes applying one or more of a lower casing algorithm, a tokenization algorithm, a punctuation mark removal algorithm, a stop word removal algorithm, a stemming algorithm, or a lemmatization algorithm.

17 . The non-transitory computer readable medium of claim 13 , wherein the similarity score is determined by applying exponential decay function to an output of the Euclidean distance formula for the set of historically similar incidents.

18 . The non-transitory computer readable medium of claim 13 , further including:

applying a normalizing algorithm to the similarity scores;

comparing the similarity scores to a threshold value;

determining a list of historically similar incidents with similarity scores greater than the threshold value; and

outputting the list of historically similar incidents to a user via a graphical user interface (GUI).

19 . The method of claim 1 , wherein the processing the data object further comprises:

extracting the multivariate information from the multiple data types; and

converting the extracted multivariate information into a concatenated string;

wherein converting the data object into a first embedding by utilizing the large language model comprises:

inputting the concatenated string into the large language model.

20 . The system of claim 7 , wherein the processing the data object further comprises:

extracting the multivariate information from the multiple data types; and

converting the extracted multivariate information into a concatenated string;

wherein converting the data object into a first embedding by utilizing the large language model comprises:

inputting the concatenated string into the large language model.

Assignments (1)

Continuity (1)

References Cited (17)