Case-based reasoning system using case generalization method

• US 9,299,025 B1
• Filed: 01/04/2013
• Issued: 03/29/2016
• Est. Priority Date: 04/06/2010
• Status: Active Grant

• Current Assignee: the united states of america as represented by the secretary of the navy
• Original Assignee: the united states of america as represented by the secretary of the navy
• Inventors: Rubin, Stuart H.
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): Wu, Fuming
• Application Number:
  US13/734,669
• Time in Patent Office: 1,180 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/025 Extracting rules from data

Abstract

A method includes comparing a user-specified context having natural language contextual antecedents to cases stored in a case base. Each stored case includes case antecedents and case consequents. A matching case exists and is selected if the case antecedents exactly match the contextual antecedents. If no match exists, a best-matching case is determined and selected. The best-matching case may be determined by generalizing the situational part of a rule and comparing the user-specified context to the stored (generalized) cases. The best-matching case is the case having the highest ratio of matching generalized case antecedents to contextual antecedents and having a matching error ratio that does not exceed an error ratio threshold. The case consequents of the selected matching case or best matching case are then displayed to a user, with case base adjustment performed based upon feedback provided by the user in response to the displayed case consequents.

Claims

1. A computer-implemented method comprising the steps of:

setting an error ratio threshold;

comparing a user-specified context comprising one or more natural language contextual antecedents to a plurality of cases stored in a case base, each case stored in the case base comprising one or more case antecedents and one or more case consequents;

determining if a matching case exists in the case base, wherein the matching case is the case having case antecedents that exactly match the contextual antecedents;

selecting the matching case if it exists, else selecting a best matching case, wherein the step of selecting a best matching case includes generalizing the case antecedents of the cases stored in the case base and comparing the user-specified context to the cases stored in the case base having generalized antecedents to determine the best-matching case, wherein the generalization of the case antecedents involves minimizing |S_i−

context|/|S_i|, where S_i is a situational variable representing a case antecedent and the context is the contextual antecedents, wherein the best matching case is the case having the highest ratio of matching case antecedents to contextual antecedents and having a matching error ratio that does not exceed the error ratio threshold; and

displaying the case consequents of the selected case to a user.

2. The computer-implemented method of claim 1, wherein if multiple best-matching cases are determined, the selected best-matching case is the most-recently-used case of the multiple best-matching cases.

3. The computer-implemented method of claim 1 further comprising the step of receiving feedback from a user regarding the displayed case consequents.

4. The computer-implemented method of claim 3, wherein if the user feedback indicates that the displayed case consequents are incorrect, the case associated with the displayed case consequents is expunged from the case base.

5. The computer-implemented method of claim 3, wherein if the user feedback indicates that the displayed case consequents are correct, case associated with the displayed case consequents is moved to the head of the case base.

6. The computer-implemented method of claim 1, wherein if a best matching case is selected, the method further comprises the step of displaying an indication to the user that a generalization was required to make a case match.

7. A non-transitory computer-readable storage medium having a method stored thereon, the method represented by computer-readable programming code, the method comprising the steps of:

setting an error ratio threshold;

comparing a user-specified context comprising one or more natural language contextual antecedents to a plurality of cases stored in a case base, each case stored in the case base comprising one or more case antecedents and one or more case consequents;

determining if a matching case exists in the case base, wherein the matching case is the case having case antecedents that exactly match the contextual antecedents;

selecting the matching case if it exists, else selecting a best matching case, wherein the step of selecting a best matching case includes generalizing the case antecedents of the cases stored in the case base and comparing the user-specified context to the cases stored in the case base having generalized antecedents to determine the best-matching case, wherein the generalization of the case antecedents involves minimizing |S_i−

context|/|S_i|, where S_i is a situational variable representing a case antecedent and the context is the contextual antecedents, wherein the best matching case is the case having the highest ratio of matching case antecedents to contextual antecedents and having a matching error ratio that does not exceed the error ratio threshold; and

displaying the case consequents of the selected case to a user.

8. The non-transitory computer-readable storage medium of claim 7, wherein if multiple best-matching cases are determined, the selected best-matching case is the most-recently-used case of the multiple best-matching cases.

9. The non-transitory computer-readable storage medium of claim 7 further comprising the step of receiving feedback from a user regarding the displayed case consequents.

10. The non-transitory computer-readable storage medium of claim 9, wherein if the user feedback indicates that the displayed case consequents are incorrect, the case associated with the displayed case consequents is expunged from the case base.

11. The non-transitory computer-readable storage medium of claim 9, wherein if the user feedback indicates that the displayed case consequents are correct, case associated with the displayed case consequents is moved to the head of the case base.

12. The non-transitory computer-readable storage medium of claim 7, wherein if a best matching case is selected, the method further comprises the step of displaying an indication to the user that a generalization was required to make a case match.

13. A system comprising:

a computer having a computer input device and a display device connected thereto; and

a plurality of distributed processors communicatively coupled to the computer wherein the computer is configured to coordinate the activities of the distributed processors, wherein each of the distributed processors is configured to maintain a case base, wherein each of the plurality of distributed processors are configured to perform a method including the steps of;

setting an error ratio threshold;

comparing a user-specified context comprising one or more natural language contextual antecedents to a plurality of cases stored in the case base, each case stored in the case base comprising one or more case antecedents and one or more case consequents;

determining if a matching case exists in the case base, wherein the matching case is the case having case antecedents that exactly match the contextual antecedents;

selecting the matching case if it exists, else selecting a best matching case, wherein the step of selecting a best matching case includes generalizing the case antecedents of the cases stored in the case base and comparing the user-specified context to the cases stored in the case base having generalized antecedents to determine the best-matching case, wherein the generalization of the case antecedents involves minimizing |S_i−

context|/|S_i|, where S_i is a situational variable representing a case antecedent and the context is the contextual antecedents, wherein the best matching case is the case having the highest ratio of matching case antecedents to contextual antecedents and having a matching error ratio that does not exceed the error ratio threshold;

displaying, via the display device, the case consequents of the selected case to a user; and

receiving, via the computer input device, feedback from the user regarding the displayed case consequent.

14. The system of claim 13, wherein if the user feedback indicates that the displayed case consequents are incorrect, the case associated with the displayed case consequents is expunged from the case base.

15. The system of claim 13, wherein if the user feedback indicates that the displayed case consequents are correct, case associated with the displayed case consequents is moved to the head of the case base.

16. The system of claim 13, wherein if a best matching case is selected, the method further comprises the step of displaying, via the display device, an indication to the user that a generalization was required to make a case match.