Method and System for Constructing a Classifier

• US 20070124300A1
• Filed: 10/18/2006
• Published: 05/31/2007
• Est. Priority Date: 10/22/2005
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bent, Graham
• Granted Patent: US 10,311,084 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

A method and system for constructing a classifier for a set of records to be classified (108) into predicted classes (351-353) are provided. The set of records that are to be classified (108) are clustered into a plurality of clusters. A first classifier (106) is created that classifies records into the plurality of clusters (321-323) and the first classifier (106) is applied to a set of training records (110), each of the training records (331-334) having a predicted class (306). A classifier (107A-C) may then be created for each sub-set of training records (341-343) classed into each of the plurality of clusters and the classifier (107A-C) applied to a sub-set of records to be classified (311-313) formed in the corresponding cluster.

Claims

1. A method of constructing a classifier for a set of records to be classified into predicted classes, comprising the steps of:

a) clustering (112) a set of records that are to be classified (108) into a plurality of clusters;

b) creating a first classifier (106) that classifies records into the plurality of clusters (321-323), and

c) applying the first classifier (106) to a set of training records (110). each of the training records having a predicted class (306).

2. A method as claimed in claim 1, wherein step c) classifies the training records into the plurality of clusters.

3. A method as claimed in claim 2, including the step of:

d) creating a classifier (107A-C) for each sub-set of training records (341-343) classed into each of the plurality of clusters.

4. A method as claimed in claim 3, including the step of:

e) applying a classifier (107A-C) created for a sub-set of training records (341-343) to a sub-set of records to be classified (311-313) formed in step a) for the corresponding cluster.

5. A method as claimed in any one of the preceding claims, wherein the records to be classified (301-304) are labelled with a cluster identifier (305) at step a).

6. A method as claimed in claim 5, wherein the records to be classified in a given sub-set (311-313) are labelled with the same cluster identifier (305).

7. A method as claimed in claim 5, wherein the training records (331-334) are labelled with a cluster identifier (305) at step c).

8. A method as claimed in claim 7, wherein the training records in a given sub-set (341-343) are labelled with the same cluster identifier (305).

9. A system for constructing a classifier for a set of records to be classified into predicted classes, comprising:

a clustering means (112) to group a set of records that are to be classified (108) into a plurality of clusters;

a first classifier (106) that classifies records into the plurality of clusters (321-323); and

a set of training records (110), each of the training records having a predicted class (306);

wherein the first classifier (106) is applied to the set of training records (110).

10. A system as claimed in claim 9, wherein the first classifier (106) classifies the training records (332-334) into the plurality of clusters.

11. A system as claimed in claim 10, including a classifier (107A-C) for each sub-set of training records (341-343) classed into each of the plurality of clusters.

12. A system as claimed in claim 11, wherein the classifier (107A-C) for a sub-set of training records (341-343) is applied to a sub-set of records to be classified (311-313) grouped in a cluster by the clustering means (112).

13. A system as claimed in any one of claims 9 to 12, wherein the clustering means (112) labels the records to be classified (301-304) with a cluster identifier (305).

14. A system as claimed in any one of claims 9 to 12, wherein the first classifier (106) labels the training records (331-334) with a cluster identifier (305).

15. A computer program product stored on a computer readable storage medium, comprising computer readable program code means for performing the steps of:

a) clustering (112) a set of records that are to be classified (108) into a plurality of clusters;

b) creating a first classifier (106) that classifies records into the plurality of clusters (321-323); and

c) applying the first classifier (106) to a set of training records (110), each of the training records having a predicted class (306).

16. A computer program product as claimed in claim 15, including the step of:

d) creating a classifier (107A-C) for each sub-set of training records (341-343) classed into each of the plurality of clusters.

17. A computer program product as claimed in claim 16, including the step of:

e) applying a classifier (107A-C) created for a sub-set of training records (341-343) to a sub-set of records to be classified (311-313) formed in step a) for the corresponding cluster.

18. A method of providing a service to a customer over a network, comprising:

a) clustering (112) a set of records that are to be classified (108) into a plurality of clusters;

b) creating a first classifier (106) that classifies records into the plurality of clusters (321-323); and

c) applying the first classifier (106) to a set of training records (110), each of the training records having a predicted class (306).

19. A method of providing a service to a customer as claimed in claim 18, including the step of:

d) creating a classifier (107A-C) for each sub-set of training records (341-343) classed into each of the plurality of clusters.

20. A method of providing a service to a customer as claimed in claim 19, including the step of:

e) applying a classifier (107A-C) created for a sub-set of training records (341-343) to a sub-set of records to be classified (311-313) formed in step a) for the corresponding cluster.