Self-healing data clusters

Patent metadata

Field	Value
Publication number	US-12254048-B2
Application number	US-202318101033-A
Country	US
Kind code	B2
Filing date	Jan 24, 2023
Priority date	Jan 24, 2023
Publication date	Mar 18, 2025
Grant date	Mar 18, 2025

Abstract

Disclosed are various embodiments for self-healing data clusters. One or more candidates are determined from the candidate pool to be evaluated with the new record. A unique pair combination is generated for each one of the candidates of the candidate pool and the new record. Next, candidate data for the one or more candidates is identified from the existing record based at least in part on one or more matching rules. A weight is assigned to one or more matching rules. Then, the candidate data of the one or more candidates and the new record is evaluated for a data linkage. A distance is calculated between each of the unique pair combinations. Finally, the candidates of the existing record and the new record are clustered into groups.

First claim

Therefore, the following is claimed:

1. A system, comprising:
a computing device comprising a processor and a memory; and
machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:
receive a transaction action from a user;
receive a data input from the user, wherein the data input is stored in a data store;
generate a new record based at least in part on the data input;
select the new record to be evaluated for a relationship with an existing record;
assign a primary key to the new record;
store the new record in the data store based at least in part on the primary key;
identify at least one candidate factor based at least in part on the data input received from the user, the candidate factor representing an account type or an entity type;
identify at least one candidate from the existing record based at least in part on the one or more candidate factors; and
assign the at least one candidate to a candidate pool, wherein the candidate pool contains at least one of the new record and the at least one candidate.

2. The system of claim 1 , wherein the machine-readable instructions further cause the computing device to evaluate candidate data of the new record and the at least one candidate in the candidate pool for a data linkage.

3. The system of claim 1 , wherein the at least one candidate is identified based at least in part on one or more matching rules.

4. The system of claim 3 , wherein the machine-readable instructions display a match result between the new record and the at least one candidate on a user interface.

5. A system, comprising:
a computing device comprising a processor and a memory; and
machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:
determine one or more candidates from a candidate pool to be evaluated for a relationship with a new record;
compute a list of unique pair combinations between each one of the one or more candidates of the candidate pool and the new record;
identify candidate data from an existing record for the one or more candidates in the candidate pool based at least in part on one or more matching rules;
evaluate candidate data between the one or more candidates of the candidate pool and the candidate of the new record for a data linkage;
determine a match between the one or more candidates and the new record based at least in part on the one or more matching rules;
calculate a distance between each of the unique pair combinations; and
cluster the candidates of the new record and the existing record into groups based on the distance.

6. The system of claim 5 , wherein the machine-readable instructions, when executed by the processor, further cause the computing device to generate a graph based at least in part on the distance between each of the unique pair combinations.

7. A system of claim 5 , wherein a score is calculated based at least in part on the distance and a primary key.

8. A system of claim 7 , wherein the machine-readable instructions that cause the computing device to calculate the score further cause the computing device to generate a model algorithm, wherein the model algorithm determines a correlation between the list of unique pair combinations based at least in part on the score.

9. The system of claim 5 , wherein the one or more candidates and the new record are assigned to a cluster based at least in part on the distance.

10. The system of claim 5 , further comprising a user interface configured to allow a user to specify the data to be analyzed for clustering.

11. The system of claim 5 , wherein the machine-readable instructions, when executed by the processor, further cause the computing device to calculate a score for each cluster.

12. A system of claim 5 , wherein the machine-readable instructions, when executed by the processor, further cause the computing device to correct at least one of an over linkage or an under linkage based at least in part on the distance.

13. A system of claim 5 , wherein the one or more candidates and the new candidate record are clustered based at least in part on the distance.

14. A system of claim 5 , wherein the match is at least one of an infinity match, an absolute match, or a partial match.

15. A method, comprising:
determining one or more candidates from a candidate pool to be evaluated for a relationship with a new candidate record;
computing a list of unique pair combinations between each one of the candidates of the candidate pool and the new candidate record;
identifying candidate data from an existing candidate record for the one or more candidates in the candidate pool based at least in part on one or more matching rules;
assigning a weight to the one or more matching rules;
evaluating candidate data between the one or more candidates and the new candidate record for a data linkage;
determine a match between the one or more candidates of the candidate pool and the candidate of the new candidate record based at least in part on the one or more matching rules;
calculating a distance between each of the unique pair combinations; and
clustering the new candidate record and the existing candidate record into groups based on the distance.

16. The method of claim 15 , further comprising calculating a score based at least in part on the distance and a primary key.

17. The method of claim 16 , further comprising correcting at least one of an over linkage or an under linkage based at least in part on the score and the distance.

18. The method of claim 15 , further comprising generating a model algorithm, wherein the model algorithm determines a correlation between the one or more candidates of the candidate pool and the candidate of new candidate record based at least in part on the distance.

19. The method of claim 15 , further comprising assigning the one or more candidates of the candidate pool and the candidate of new candidate record to a cluster based at least in part on the distance.

20. The method of claim 15 , wherein the match is at least one of an infinity match, an absolute match, or a partial match.

Assignees

• American Express Travel Related Services Co Inc

Inventors

• Silverstein Karina I Alvarez
• Dhawan Mahasweta
• D'Souza Alan Wilson
• Dutta Sayantan
• Handa Anmol
• Jangra Shipali
• Kamil Hen Israel
• Modadugu Rajendra Prasad
• Prajapati Sumit
• SHARMA Bhupesh

Classifications

• G06F16/906 Primary

  Clustering; Classification · CPC title