Method and apparatus for determining rule in database

• US 6,317,735 B1
• Filed: 05/28/1999
• Issued: 11/13/2001
• Est. Priority Date: 05/29/1998
• Status: Expired due to Term

• Current Assignee: SAP SE
• Original Assignee: International Business Machines Corporation
• Inventors: Morimoto, Yasuhiko
• Primary Examiner(s): Choules, Jack
• Assistant Examiner(s): RAYYAN, SUSAN F
• Application Number:
  US09/322,215
• Time in Patent Office: 900 Days
• Field of Search: 707/2, 707/1
• US Class Current: 707/694
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06F 18/245 relating to the decision surface
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: Y10S 707/955 Object-oriented
• CPC Class Codes: Y10S 707/962 Entity-attribute-value
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing

Abstract

When determining a rule associated with an objective attribute of data in a database to predict the objective attribute value of data, following steps are performed: storing values relative to data belonging to each bucket wherein the values correspond to each bucket in a plane, and the plane has two axes respectively corresponding to two predicative numeric attributes of data and is divided into N×M buckets; segmenting a bucket region that is satisfied with a predetermined condition, from the plane; performing a smoothing processing for a boundary of the segmented bucket region; and determining a rule for predicting the object attribute value of the data by the smoothed region.

Claims

1. A method for determining a rule associated with an objective attribute of data in a database to predict the objective attribute value of data, said method comprising the steps of:

storing values relative to data belonging to each bucket, said values corresponding to each said bucket in a plane, wherein said plane has two axes respectively corresponding to two predicative numeric attributes of data and is divided into N×

M buckets;

segmenting a bucket region that is satisfied with a predetermined condition, from said plane;

performing a smoothing processing for a boundary of the segmented bucket region; and

determining a rule for predicting the object attribute value of said data by the smoothed region.

2. The method set forth in claim 1, wherein said values relative to data belonging to a bucket are the number of data belonging to each bucket and values relative to said objective attribute of data belonging to a bucket.

3. The method set forth in claim 2, wherein said objective attribute is a numeric attribute, and said value relative to said objective attribute is the sum of values of said objective attribute of data belonging to a bucket.

4. The method set forth in claim 2, wherein said objective attribute is a true-false attribute, and said value relative to said objective attribute is the number of data whose objective attribute value is true and belongs to a bucket.

5. The method set forth in claim 1, wherein said predetermined condition is to maximize an interclass variance.

6. The method set forth in claim 1, wherein said predetermined condition is to maximize an entropy gain of discrete values of said objective attribute.

7. The method set forth in claim 1, wherein said predetermined condition is to maximize a GINI index function value of discrete values of said objective attribute.

8. The method set forth in claim 1, wherein said predetermined condition is to maximize a χ

square value of discrete values of said objective attribute.

9. The method set forth in claim 1, wherein said objective attribute is a true-false attribute, and said predetermined condition is to maximize the number of included data when a ratio of data whose objective attribute value is true is more than a predetermined value, or to maximize a ratio of data whose objective attribute value is true when the minimum number of included data is defined.

10. The method set forth in claim 1, wherein said smoothing processing is a processing to make a boundary of said region a spline curve.

11. The method set forth in claim 1, wherein said smoothing processing comprises a step of defining control points in sides of N stripes in said region, wherein said stripes are parallel to an axis corresponding to a first predicative numeric attribute, and said sides are parallel to an axis corresponding to a second predicative numeric attribute.

12. The method set forth in claim 11, wherein said smoothing processing further comprises a step of setting a curve passing said control points.

13. The method set forth in claim 1, wherein said determining step comprises a step of generating a tree in which a node relative to data within the smoothed region and a node relative to data outside of the smoothed region are added.

14. An apparatus for determining a rule associated with an objective attribute of data in a database to predict the objective attribute value of data, comprising:

means for storing values relative to data belonging to each bucket, said values corresponding to each said bucket in a plane, wherein said plane has two axes respectively corresponding to two predicative numeric attributes of data and is divided into N×

M buckets;

means for segmenting a bucket region that is satisfied with a predetermined condition, from said plane;

means for performing a smoothing processing for a boundary of the segmented bucket region; and

means for determining a rule for predicting the object attribute value of said data by the smoothed region.

15. An apparatus for determining a rule associated with an objective attribute of data in a database to predict the objective attribute value of data, said apparatus comprising:

a means for storing values relative to data belonging to each bucket, said values corresponding to each said bucket in a plane, wherein said plane has two axes respectively corresponding to two predicative numeric attributes of data and is divided into N×

M buckets;

a means for segmenting a bucket region that is satisfied with a predetermined condition, from said plane;

a means for performing a smoothing processing for a boundary of the segmented bucket region; and

a means for determining a rule for predicting the object attribute value of said data by the smoothed region.

16. A storage medium for storing a program for causing a computer to determine a rule associated with an objective attribute of data in a database to predict the objective attribute value of data, said program comprising the steps of:

storing values relative to data belonging to each bucket, said values corresponding to each said bucket in a plane, wherein said plane has two axes respectively corresponding to two predicative numeric attributes of data and is divided into N×

M buckets;

segmenting a bucket region that is satisfied with a predetermined condition, from said plane;

performing a smoothing processing for a boundary of the segmented bucket region; and

determining a rule for predicting the object attribute value of said data by the smoothed region.

17. The storage medium set forth in claim 16, wherein said values relative to data belonging to a bucket are the number of data belonging to each bucket and values relative to said objective attribute of data belonging to a bucket.

18. The storage medium set forth in claim 16, wherein said smoothing processing is a processing to make a boundary of said region a spline curve.

19. The storage medium set forth in claim 16, wherein said determining step comprises a step of generating a tree in which a node relative to data within the smoothed region and a node relative to data outside of the smoothed region are added.