Machine learning based enrichment of database objects

• US 10,509,782 B2
• Filed: 12/11/2017
• Issued: 12/17/2019
• Est. Priority Date: 12/11/2017
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP SE
• Inventors: Banerjee, Debashis, Rajagopalan, Nithya
• Primary Examiner(s): Colan, Giovanna B
• Application Number:
  US15/838,256
• Publication Number:
  US 20190179937A1
• Time in Patent Office: 736 Days
• CPC Class Codes: G06F 16/2282 Tablespace storage structures; Management thereof
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

A method for enriching an object in a database may include determining, by a trained machine learning model, that a first object at the database is same and/or similar to a second object at the database. The first object and the second object may be part of a schema of the database. The second object may be subordinate to the first object. In response to the determination that the first object is same and/or similar to the second object, one or more attributes associated with the second object may be added to the first object. Related systems and articles of manufacture including computer program products are also provided.

Claims

1. A system, comprising:

at least one data processor; and

at least one memory storing instructions which, when executed by the at least one data processor, cause operations comprising;

determining, by a trained machine learning model, that a first object at a database is same and/or similar to a second object at the database, the first object and the second object comprising a schema of the database, and the second object being subordinate to the first object; and

in response to the determination that the first object is same and/or similar to the second object, adding, to the first object, one or more attributes associated with the second object.

2. The system of claim 1, wherein the trained machine learning model is configured to generate at least one cluster of objects comprising a grouping of same and/or similar objects at the database, and wherein the first object is determined to be same and/or similar to the second object based at least on the first object and the second object both being in a same cluster of objects.

3. The system of claim 1, wherein the trained machine learning model is configured to generate a first cluster of objects comprising a grouping of same and/or similar objects at the database, wherein the first cluster of objects includes the second object, and wherein the first object is determined to be same and/or similar to the second object based at least on a distance between the first object and the first cluster of objects.

4. The system of claim 3, wherein the trained machine learning model further generates a second cluster of objects comprising another grouping of same and/or similar objects at the database, and wherein the first object is determined to be same and/or similar to the second object based at least on the distance between the first object and the first cluster of objects being less than a distance between the first object and the second cluster of objects.

5. The system of claim 1, wherein the second object descends directly and/or indirectly from the first object such that the second object inherits at least one attribute associated with the first object.

6. The system of claim 1, wherein the one or more attributes are added to the first object by at least adding, to a database table corresponding to the first object, at least one column for the one or more attributes.

7. The system of claim 1, wherein the one or more attributes are added to the first object by at least adding, to a structured data representation of the first object, at least one key-value corresponding to the one or more attributes.

8. The system of claim 1, further comprising:

training, based at least on training data, a machine learning model to identify same and/or similar objects in the training data.

9. The system of claim 8, wherein the training data comprises one or more existing objects and/or existing schemas that include objects known to be same, similar, and/or different.

10. The system of claim 1, wherein the trained machine learning model comprises a cluster analyzer, a neural network, a support vector machine, a scorecard, logistic regression model, a Bayesian model, and/or a decision tree.

11. A computer-implemented method, comprising:

determining, by a trained machine learning model, that a first object at a database is same and/or similar to a second object at the database, the first object and the second object comprising a schema of the database, and the second object being subordinate to the first object; and

in response to the determination that the first object is same and/or similar to the second object, adding, to the first object, one or more attributes associated with the second object.

12. The method of claim 11, wherein the trained machine learning model is configured to generate at least one cluster of objects comprising a grouping of same and/or similar objects at the database, and wherein the first object is determined to be same and/or similar to the second object based at least on the first object and the second object both being in a same cluster of objects.

13. The method of claim 11, wherein the trained machine learning model is configured to generate a first cluster of objects comprising a grouping of same and/or similar objects at the database, wherein the first cluster of objects includes the second object, and wherein the first object is determined to be same and/or similar to the second object based at least on a distance between the first object and the first cluster of objects.

14. The method of claim 13, wherein the trained machine learning model further generates a second cluster of objects comprising another grouping of same and/or similar objects at the database, and wherein the first object is determined to be same and/or similar to the second object based at least on the distance between the first object and the first cluster of objects being less than a distance between the first object and the second cluster of objects.

15. The method of claim 11, wherein the second object descends directly and/or indirectly from the first object such that the second object inherits at least one attribute associated with the first object.

16. The method of claim 11, wherein the one or more attributes are added to the first object by at least adding, to a database table corresponding to the first object, at least one column for the one or more attributes.

17. The method of claim 11, wherein the one or more attributes are added to the first object by at least adding, to a structured data representation of the first object, at least one key-value corresponding to the one or more attributes.

18. The method of claim 11, further comprising:

training, based at least on training data, a machine learning model to identify same and/or similar objects in the training data.

19. The method of claim 18, wherein the training data comprises one or more existing objects and/or existing schemas that include objects known to be same, similar, and/or different.

20. A non-transitory computer-readable medium including instructions, which when executed by at least one data processor, cause operations comprising:

determining, by a trained machine learning model, that a first object at a database is same and/or similar to a second object at the database, the first object and the second object comprising a schema of the database, and the second object being subordinate to the first object; and

in response to the determination that the first object is same and/or similar to the second object, adding, to the first object, one or more attributes associated with the second object.