Machine learning-based rule mining algorithm

Patent metadata

Field	Value
Publication number	US-11783205-B2
Application number	US-201916717819-A
Country	US
Kind code	B2
Filing date	Dec 17, 2019
Priority date	Dec 17, 2019
Publication date	Oct 10, 2023
Grant date	Oct 10, 2023

Abstract

Data is received that defines a rule mining run including a scope of a search and at least one data source to be searched. In response, the at least one data source is polled to obtain rules responsive to the rule mining run. Each rule can specify one or more actions to take as part of a computer-implemented process when certain conditions are met. A list of rules (i.e., a proposed subset of the obtained rules) can then be generated using at least one machine learning model. The generated list of rule can then be displayed in a graphical user interface. Related apparatus, systems, techniques and articles are also described.

First claim

What is claimed is:

1. A computer-implemented method comprising;
receiving data defining a rule mining run including a mining scope and at least one data source to be mined;
polling the at least one data source to obtain rules responsive to the rule mining run, the rules each specifying one or more actions to take as part of a computer-implemented process when certain conditions are met;
generating, using at least one machine learning model comprising a K-Optimal Rule Discovery (KORD) algorithm, a list of rules from the obtained rules and information to help make a decision of whether or not to implement a given rule out of the list of rules, wherein the information comprises how much data from the mining scope matches the given rule, how much data from the mining scope violates the given rule, and how much data of the mining scope is not relevant to the given rule; and
causing the generated list of rules to be displayed in a graphical user interface, wherein the graphical user interface indicates how much data from the mining scope matches the given rule, how much data from the mining scope violates the given rule, and how much data of the mining scope is not relevant to the given rule.

2. The method of claim 1 , wherein the at least one machine learning model uses association rules to identify rules most relevant to the mining scope.

3. The method of claim 1 , wherein the at least one machine learning model optimizes a user-specified measure of rule value with respect to a set of sample data and user-specified constraints.

4. The method of claim 3 , wherein the K-Optimal Rule Discovery (KORD) algorithm calculates support, confidence, leverage, or lift.

5. The method of claim 1 , wherein the graphical user interface comprises graphical user interface elements allowing for sorting of the list of rules and for accepting or rejecting rules.

6. The method of claim 5 , wherein the rules are business rules and the computer-implemented process is a computer-implemented business process.

7. The method of claim 5 further comprising:
activating rules via the graphical user interface in a rules repository utilized by the computer-implemented processes.

8. A system comprising;
at least one data processor; and
memory including instructions which, when executed by the at least one data processor, result in operations comprising:
receiving data defining a rule mining run including a mining scope and at least one data source to be mined;
polling the at least one data source to obtain rules responsive to the rule mining run, the rules each specifying one or more actions to take as part of a computer-implemented process when certain conditions are met;
generating, using at least one machine learning model comprising a K-Optimal Rule Discovery (KORD) algorithm, a list of rules from the obtained rules and information to help make a decision of whether or not to implement a given rule out of the list of rules, wherein the information comprises how much data from the mining scope matches the given rule, how much data from the mining scope violates the given rule, and how much data of the mining scope is not relevant to the given rule; and
causing the generated list of rules to be displayed in a graphical user interface, wherein the graphical user interface comprises a plurality of graphical user interface elements that visually indicate what percentage of data from the mining scope matches the given rule, what percentage of data from the mining scope violates the given rule, and what percentage of data of the mining scope is not relevant to the given rule.

9. The system of claim 8 , wherein the at least one machine learning model uses association rules to identify rules most relevant to the mining scope.

10. The system of claim 8 , wherein the at least one machine learning model optimizes a user-specified measure of rule value with respect to a set of sample data and user-specified constraints.

11. The system of claim 8 , wherein the graphical user interface comprises graphical user interface elements allowing for sorting of the list of rules and for accepting or rejecting rules.

12. The system of claim 11 , wherein the rules are business rules and the computer-implemented process is a computer-implemented business process.

13. The system of claim 11 , wherein the operations further comprise:
activating rules via the graphical user interface in a rules repository utilized by the computer-implemented processes.

14. A non-transitory computer program product including instructions which, when executed by at least one computing device, result in operations comprising:
receiving data defining a rule mining run including a mining scope and at least one data source to be mined;
polling the at least one data source to obtain rules responsive to the rule mining run, the rules each specifying one or more actions to take as part of a computer-implemented process when certain conditions are met;
generating, using at least one machine learning model comprising a K-Optimal Rule Discovery (KORD) algorithm, a list of rules forming a subset of the obtained rules and information to help make a decision of whether or not to implement a given rule out of the list of rules, wherein the information comprises how much data from the mining scope matches the given rule, how much data from the mining scope violates the given rule, and how much data of the mining scope is not relevant to the given rule; and
causing the generated list of rules to be displayed in a graphical user interface, wherein the graphical user interface comprises a plurality of graphical user interface elements that visually indicate what percentage of data from the mining scope matches the given rule, what percentage of data from the mining scope violates the given rule, and what percentage of data of the mining scope is not relevant to the given rule.

15. The computer program product of claim 14 , wherein the at least one machine learning model optimizes a user-specified measure of rule value with respect to a set of sample data and user-specified constraints.

16. The computer program product of claim 14 , wherein the graphical user interface comprises graphical user interface elements allowing for sorting of the list of rules and for accepting or rejecting rules.

17. The computer program product of claim 16 , wherein the rules are business rules and the computer-implemented process is a computer-implemented business process.

18. The computer program product of claim 17 , wherein the operations further comprise:
activating rules via the graphical user interface in a rules repository utilized by the computer-implemented processes.

Assignees

• Sap Se

Inventors

• Wang Kefeng
• Seifried Andreas
• Bruegel Birgitta
• Turley Kieran
• Raev Dimitrij

Classifications

• G06N5/025 Primary

  Extracting rules from data · CPC title

• G06N20/00

  Machine learning · CPC title