METHOD AND APPARATUS FOR OBTAINING CONSTRAINTS ON EVENTS

• US 20160147634A1
• Filed: 11/20/2015
• Published: 05/26/2016
• Est. Priority Date: 11/21/2014
• Status: Abandoned Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Bing, Li, Xiang, LIAN, XIAO JIAN, Liu, Dan, Liu, Haifeng, Mei, Jing, Xie, Guo Tong, Yu, Yi Qin, Zhang, Jing
• Application Number:
  US14/946,815
• Publication Number:
  US 20160147634A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 11/3608 using formal methods, e.g. model checking, abstract interpretation theorem proving G06N5/013
• CPC Class Codes: G06F 11/3692 for test results analysis

Abstract

Method and apparatus for obtaining constraints on events. The method includes: obtaining a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event, wherein each event sequence among the multiple event sequences is a series of historical events that are executed for achieving the goal; identifying an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence; and in response to metric of the impact satisfying a predefined condition, determining the at least one part of candidate constraints as the constraint. An apparatus for determining a constraint on an event and a method and apparatus for generating a Case Management Model from multiple event sequences are also provided.

Claims

1. A computer-implemented method for determining a constraint on an event, the method comprising:

obtaining a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event, wherein each event sequence among the multiple event sequences is a series of historical events that are executed for achieving the goal;

identifying an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence; and

determining the at least one part of candidate constraints as the constraint in response to a metric of the impact satisfying a predefined condition.

2. The computer-implemented method according to claim 1, wherein the step of obtaining a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event comprises:

constructing a constraint feature vector associated with each occurrence, wherein each element in the constraint feature vector represents the status of one candidate constraint among the multiple candidate constraints; and

acquiring the correspondence based on the constraint feature vector and the goal.

3. The computer-implemented method according to claim 2, wherein the step of constructing a constraint feature vector associated with the each occurrence comprises:

determining the status of each candidate constraint among the multiple candidate constraints in an event sequence associated with the each occurrence with respect to each element in the constraint feature vector; and

building the constraint feature vector based on the status of the each candidate constraint with respect to each element in the constraint feature vector.

4. The computer-implemented method according to claim 1, wherein the step of identifying an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence comprises:

generating at least one subset of the multiple candidate constraints; and

identifying the impact on the goal of a candidate constraint in the each subset based on the correspondence.

5. The computer-implemented method according to claim 4, wherein the step of generating at least one subset of the multiple candidate constraints comprises:

generating a 1-item subset of the multiple candidate constraints, wherein each element in the 1-item subset comprises one candidate constraint among the multiple candidate constraints.

6. The computer-implemented method according to claim 5, further comprising:

generating an n-item subset of the multiple candidate constraints based on an (n−

1)-item subset of the multiple candidate constraints, wherein 2≦

n≦

the number of the multiple candidate constraints, and each element in the n-item subset comprises n candidate constraints.

7. The computer-implemented method according to claim 4, wherein the step of identifying the impact on the goal of a candidate constraint in the each subset based on the correspondence comprises:

determining correlation on the goal of the candidate constraint in the each subset as the impact based on the correspondence.

8. The computer-implemented method according to claim 7, further comprising:

determining support of the candidate constraint in the each subset based on the correspondence; and

adding the support to the impact.

9. The computer-implemented method according to claim 1, wherein types of the multiple candidate constraints comprise at least one of:

existence constraint, temporal constraint, and data constraint.

10. A computer-implemented method for generating a Case Management Model from multiple event sequences, the method comprising:

extracting multiple events from the multiple event sequences;

obtaining a constraint on each event; and

generating the Case Management Model based on the each event and the constraint on the each event.

11. The computer-implemented method according to claim 10, wherein the step of obtaining a constraint on each event comprises:

obtaining a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event, wherein each event sequence among the multiple event sequences is a series of historical events that are executed for achieving the goal;

identifying an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence; and

determining the at least one part of candidate constraints as the constraint in response to a metric of the impact satisfying a predefined condition.

12. An apparatus for determining a constraint on an event, the apparatus comprising:

an obtaining module configured to obtain a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event, wherein each event sequence among the multiple event sequences is a series of historical events that are executed for achieving the goal;

an identifying module configured to identify an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence; and

a determining module configured to determine the at least one part of candidate constraints as the constraint in response to metric of the impact satisfying a predefined condition.

13. The apparatus according to claim 12, wherein the obtaining module comprises:

a constructing module configured to construct a constraint feature vector associated with the each occurrence, wherein each element in the constraint feature vector represents the status of one candidate constraint among the multiple candidate constraints; and

an acquiring module configured to acquire the correspondence based on the constraint feature vector and the goal.

14. The apparatus according to claim 13, wherein the constructing module further comprises:

a status determining module configured to determine status of each candidate constraint among the multiple candidate constraints in an event sequence associated with the each occurrence; and

a building module configured to build the constraint feature vector based on the status of the each candidate constraint.

15. The apparatus according to claim 12, wherein the identifying module comprises:

a generating module configured to generate at least one subset of the multiple candidate constraints; and

an impact identifying module configured to identify the impact on the goal of a candidate constraint in the each subset based on the correspondence.

16. The apparatus according to claim 15, wherein the generating module comprises:

a first subset generating module configured to generate a 1-item subset of the multiple candidate constraints, wherein each element in the 1-item subset comprises one candidate constraint among the multiple candidate constraints.

17. The apparatus according to claim 16, further comprising:

a second subset generating module configured to generate an n-item subset of the multiple candidate constraints based on an (n−

1)-item subset of the multiple candidate constraints, wherein 2≦

n≦

the number of the multiple candidate constraints, and each element in the n-item subset comprises n candidate constraints.

18. The apparatus according to claim 15, wherein the impact identifying module comprises:

a correlation calculating module configured to determine correlation on the goal of the candidate constraint in the each subset as the impact based on the correspondence.

19. The apparatus according to claim 18, further comprising:

a support calculating module configured to calculate support of the candidate constraint in the each subset based on the correspondence; and

an adding module configured to add the support to the impact.

20. The apparatus according claim 12, wherein types of the multiple candidate constraints comprise at least one of:

existence constraint, temporal constraint, and data constraint.

21. An apparatus for generating a Case Management Model from multiple event sequences, the apparatus comprising:

an extracting module configured to extract multiple events from the multiple event sequences;

an obtaining module configured to obtain a constraint on the each event with respect to each event among the multiple events; and

a generating module configured to generate the Case Management Model based on the each event and the constraint on the each event.

22. The apparatus according to claim 21, wherein the obtaining module comprises:

an obtaining module configured to obtain a correspondence between a goal and multiple candidate constraints associated with the event from multiple event sequences including the event, wherein each event sequence among the multiple event sequences is a series of historical events that are executed for achieving the goal;

an identifying module configured to identify an impact on the goal of at least one part of candidate constraints among the multiple candidate constraints based on the correspondence; and

a determining module configured to determine the at least one part of candidate constraints as the constraint in response to metric of the impact satisfying a predefined condition.