Providing information regarding a trend based on output of a categorizer

• US 7,437,338 B1
• Filed: 03/21/2006
• Issued: 10/14/2008
• Est. Priority Date: 03/21/2006
• Status: Active Grant

• Current Assignee: Valtrus Innovations Limited (f/k/a Dolya Holdco 9 Limited) (Key Patent Innovations Limited)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Kirshenbaum, Evan R., Suermondt, Henri J., Forman, George H.
• Primary Examiner(s): Holmes; Michael B
• Application Number:
  US11/386,005
• Time in Patent Office: 938 Days
• Field of Search: None
• US Class Current: 706/20
• CPC Class Codes: G06F 16/2455 Query execution
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

In response to a query relating to a category, identify at least one case. An indication is received regarding whether the at least one case belongs to the category. A categorizer is trained based on the received indication. Cases from a data set are allocated into plural bins. For respective bins, quantification measures are computed for cases belonging to the category based on output from the categorizer. Information is provided regarding a trend of the category over the plural bins based on the computed quantification measures.

Claims

1. A method executed by a computer, comprising:

identifying, in response to a query relating to a category, at least one case;

receiving an indication confirming or disconfirming that the at least one case belongs to the category;

training a categorizer based on the received indication;

allocating cases from a data set into plural bins; and

for respective bins, computing quantification measures for cases belonging to the category based on output from the categorizer; and

providing information regarding a first trend of the category over the plural bins based on the computed quantification measures.

2. The method of claim 1, wherein the plural bins comprise one of time points and time windows, the method further comprising providing a time-based graph of the quantification measures over the one of the time points and time windows to provide information regarding the first trend.

3. The method of claim 1, further comprising identifying, based on the first trend, that an issue associated with the category is changing.

4. The method of claim 3, further comprising:

after an action is taken with respect to an issue associated with the category in response to determining that the issue associated with the category is changing based on the first trend,

receiving additional cases,

allocating the additional cases into additional bins,

computing quantification measures in respective additional bins for the additional cases belonging to the category based on output from the categorizer, and

presenting information regarding a second trend of the category over the additional bins based on the computed quantification measures in the respective additional bins.

5. The method of claim 4, further comprising determining, based on a comparison of the first trend with the second trend, an effect of the action with respect to the category.

6. The method of claim 1, wherein computing the quantification measures is performed by a quantifier in cooperation with the categorizer.

7. The method of claim 1, wherein allocating cases from the data set into the plural bins comprises allocating cases from the data set into bins based on at least one of a time criterion, a product number criterion, a product type criterion, a product age criterion, a size criterion, a speed criterion, a cost criterion, a warranty type criterion, a customer type criterion, a customer interaction type criterion, a location criterion, a usage criterion, a serial number criterion, an altitude criterion, a temperature criterion, a humidity criterion, a pressure criterion, a customer income level criterion, and a purchase price criterion.

8. The method of claim 7, further comprising ordering the plural bins according to the at least one of the time criterion, the product number criterion, the product type criterion, the product age criterion, the size criterion, the speed criterion, the cost criterion, the warranty type criterion, the customer type criterion, the customer interaction type criterion, the location criterion, the usage criterion, the serial number criterion, the altitude criterion, the temperature criterion, the humidity criterion, the pressure criterion, the customer income level criterion, and the purchase price criterion.

9. The method of claim 1, wherein providing the information regarding the first trend comprises providing the information using one or plural graphs of the quantification measures.

10. The method of claim 1, further comprising detecting a systematic change in the category using a change-point detection algorithm applied to the quantification measures.

11. The method of claim 10, wherein detecting the systematic change using the change-point detection algorithm comprises using one of a cumulative sums algorithm and a generalized likelihood ratio algorithm.

12. The method of claim 1, further comprising creating a model based on the first trend to predict future values for the quantification measures.

13. The method of claim 1, wherein allocating cases from the data set into the plural bins comprises allocating cases from the data set into sliding time windows.

14. The method of claim 1, further comprising:

computing additional quantification measures in respective bins for cases belonging to additional categories based on further output from the categorizer;

providing information regarding additional trends of respective additional categories over the plural bins based on the computed additional quantification measures.

15. The method of claim 14, further comprising providing information to enable ordering of the categories based on the trends.

16. The method of claim 15, wherein providing information to enable the ordering of the categories is based on at least one of:

(1) relative rates of change of the trends of respective categories, (2) presence or lack thereof of sharp changes in the trends of respective categories, and (3) relative magnitudes of projected future quantification measures of respective categories based on the trends.

21. The method of claim 1, wherein receiving the indication confirming or disconfirming comprises receiving the indication of user confirmation or disconfirmation of whether the at least one case belongs to the category.

17. A method executed by a computer comprising:

identifying at least one case in response to a query relating to a category;

receiving an indication confirming or disconfirming that the identified at least one case belongs to the category;

training the categorizer based on the received indication;

classifying, with the trained categorizer, cases from a data set into the category; and

allocating the cases into plural bins;

computing quantification measures for cases in respective bins;

providing information regarding a trend of the category over the plural bins based on the computed quantification measures; and

determining whether an action take with respect to the category has resulted in a change of the quantification measures.

18. The method of claim 17, further comprising developing training cases based on the received indication, wherein the categorizer is trained based on the training cases.

19. The method of claim 17, wherein the trend represents an increasing or decreasing direction of the quantification measures.

22. The method of claim 17, wherein receiving the indication comprises receiving the indication of user confirmation or disconfirmation of whether the at least one case belongs to the category.

20. A computer-readable storage medium storing instructions that when executed cause a computer to:

identify, based on a query relating to a category, at least one case;

receive an indication confirming or disconfirming that the identified at least one case belongs to the category;

train a categorizer based on the received indication;

classify, by the trained categorizer, cases from a data set into the category;

divide the cases into plural bins;

compute quantification measures for the cases in respective bins; and

display a representation of a trend of the category over the bins based on the computed quantification measures.

23. The computer-readable storage medium of claim 20, wherein receiving the indication confirming or disconfirming comprises receiving the indication of user confirmation or disconfirmation of whether the at least one case belongs to the category.