Producing a measure regarding cases associated with an issue after one or more events have occurred

• US 8,719,073 B1
• Filed: 08/25/2005
• Issued: 05/06/2014
• Est. Priority Date: 08/25/2005
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Kirshenbaum, Evan R., Suermondt, Henri J., Forman, George H.
• Primary Examiner(s): Singh, Gurkanwaljit
• Application Number:
  US11/211,979
• Time in Patent Office: 3,176 Days
• US Class Current: 705/7.32
• CPC Class Codes: G06F 16/24542 Plan optimisation
• CPC Class Codes: G06N 5/048 Fuzzy inferencing
• CPC Class Codes: G06Q 30/016 After-sales
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

A categorizer produces a first measure regarding cases associated with an issue. Information regarding additional cases associated with the issue is received after one or more events have occurred with respect to the issue. Based on further output from the categorizer, a second measure is produced regarding the additional cases associated with the issue.

Claims

1. A method comprising:

producing, based on output from a categorizer executed on one or more processors, a first measure regarding cases associated with an issue;

receiving information regarding additional cases associated with the issue after one or more events have occurred with respect to the issue;

after the one or more events have occurred, producing, based on further output from the categorizer executed on the one or more processors, a second measure regarding the additional cases associated with the issue; and

training the categorizer, wherein training the categorizer comprises;

receiving a query relating to the issue;

identifying cases in response to the query;

receiving user confirmation or disconfirmation that the identified cases belong to the issue; and

developing training cases for training the categorizer based on receiving the user confirmation or disconfirmation.

2. The method of claim 1, further comprising comparing the first measure with the second measure to determine an impact of the one or more events with respect to the issue.

3. The method of claim 2, further comprising providing an indication of the comparing of the first and second measures.

4. The method of claim 3, wherein providing the indication comprises generating a report containing information relating to the comparing of the first and second measures.

5. The method of claim 1, wherein producing the first and second measures regarding the cases associated with the issue comprises producing the first and second measures regarding the cases associated with a problem, and

wherein receiving the information after the one or more events have occurred comprises receiving the information after a fix has been made available for the problem.

6. The method of claim 1, wherein producing the first and second measures regarding cases associated with the issue comprises producing the first and second measures regarding cases associated with an issue faced by a first organization, and

wherein receiving information after the one or more events have occurred comprises receiving the information after an event associated with a second organization has occurred.

7. The method of claim 1, wherein producing the first and second measures is performed by a quantifier associated with the categorizer.

8. The method of claim 1, wherein producing the first measure comprises:

computing an actual measure; and

adjusting, based on at least one factor, the actual measure to produce the first measure.

9. The method of claim 8, wherein adjusting the actual measure comprises extrapolating the measure for a future time period.

10. The method of claim 8, wherein adjusting the actual measure comprises normalizing the actual measure based on the at least one factor including at least one of seasonality, sales volume, product end-of-life, marketing, installed base under support contract, and installed base under warranty.

11. The method of claim 8, wherein producing the second measure comprises:

computing a second actual measure; and

adjusting, based on at least one factor, the second actual measure to produce the second measure.

12. The method of claim 11, further comprising comparing the first and second measures to determine an impact of the one or more events with respect to the issue.

13. The method of claim 1, wherein the one or more events comprise one or more fixes for the issue, the method further comprising:

comparing the first and second measures;

receiving cost information relating to the one or more fixes; and

reporting a return on investment based on comparing the first and second measures and the cost information.

14. The method of claim 1, further comprising identifying the one or more events for the issue, the one or more events comprising at least one of changing a product design;

creating a software patch;

creating a software program;

creating printed documentation;

altering documentation;

creating on-line documentation;

identifying existing documentation;

altering availability of a document;

creating automated diagnostic software;

providing documentation to customer support representatives;

identifying an action by a third party;

allocating a sum of money for the issue;

allocating a physical resource for the issue;

hiring a person;

assigning a task to a person;

altering an organization of a web site;

modifying a packaging of a product;

modifying a manufacturing process for a product;

contacting a customer, vendor, supplier, employee, or partner;

modifying a marketing campaign;

changing response time of service providers;

training service personnel;

discontinuing efforts that are no longer required;

changing the process of writing and delivery of software programs;

taking actions with reference to seasonal fluctuations;

providing reports to customers regarding how the issue is being monitored and addressed;

recalling a product;

making available a repair for a product;

making available a modification for a product;

changing recommended usage of a product;

release of a new operating system by a third party vendor;

release of a patch from a third party software vendor;

release of a new hardware platform by a third party vendor; and

new governmental regulation.

15. The method of claim 1, wherein producing the first measure comprises producing at least one of a measure of a duration of calls;

cost of repair;

amount of time to resolve the calls;

amount of lost revenue;

an amount of money to resolve calls;

amount of lost revenue due to calls;

a degree of customer aggravation;

an amount of time spent by customers before initiating a call;

an amount of time spent by customers during the call;

an amount of time spent diagnosing cases;

an amount of money spent by customers;

an amount of money spent interacting with customers;

an amount of money spent resolving the cases;

a number of customers who declined to pay for a call;

an amount estimating expected lost future revenue from the issue.

16. The method of claim 1, wherein the issue comprises a first problem, the method further comprising:

receiving a list of problems; and

selecting the first problem from among the list of problems.

17. The method of claim 16, further comprising receiving information prioritizing the problems in the list.

18. The method of claim 16, wherein selecting the first problem from among the list of problems is based on a prediction of a magnitude of further deterioration if the first problem is not addressed as compared to amount of savings expected from addressing the first problem.

19. The method of claim 1, further comprising generating a probability distribution of cases associated with the issue, the probability distribution containing weights associated with respective cases,

wherein producing the first measure is based on the weights.

32. The method of claim 1, wherein producing the first measure and producing the second measure are performed based on output of the trained categorizer.

33. The method of claim 32, wherein developing the training cases comprises:

adding at least one of the identified cases to a positive training set in response to receiving user confirmation; and

adding at least one of the identified cases to a negative training set in response to receiving user disconfirmation,

wherein the developed training cases include the positive training set and the negative training set.

20. A system, comprising:

one or more processors;

a categorizer executable on the one or more processors for categorizing cases in a data set into at least a first category, wherein the categorizer is a trained categorizer trained using a technique comprising;

receiving a query relating to the first category;

identifying cases in response to the query;

receiving user confirmation or disconfirmation that the identified cases belong to the first category; and

developing training cases for training the categorizer based on receiving the user confirmation or disconfirmation;

a quantifier executable on the one or more processors in cooperation with the trained categorizer to compute a first quantification measure for a first group of cases, and a second quantification measure for a second group of cases, wherein the first group of cases relates to cases prior to occurrence of an event corresponding to an issue associated with the first category, and the second group of cases relates to cases after occurrence of the event; and

a module executable on the one or more processors to determine an impact of the event on the issue based on comparing the first and second quantification measures.

21. The system of claim 20, wherein the quantifier is executable on the one or more processors to produce the first quantification measure by adjusting a first actual measure based on at least one factor,

wherein adjusting the first actual measure comprises one of extrapolating and normalizing the first actual measure.

34. The system of claim 20, wherein developing the training cases comprises:

adding at least one of the identified cases to a positive training set in response to receiving user confirmation; and

adding at least one of the identified cases to a negative training set in response to receiving user disconfirmation,

wherein the developed training cases include the positive training set and the negative training set.

22. A computer-readable storage medium storing instructions that when executed cause a system having a processor to:

produce, based on output from a categorizer, a first measure regarding cases associated with an issue;

receive information regarding additional cases associated with the issue after one or more events have occurred with respect to the issue;

produce, after the one or more events have occurred and based on further output from the categorizer, a second measure regarding the additional cases associated with the issue; and

train the categorizer, wherein training the categorizer comprises;

receiving a query relating to the issue;

identifying cases in response to the query;

receiving user confirmation or disconfirmation that the identified cases belong to the issue; and

developing training cases for training the categorizer based on receiving the user confirmation or disconfirmation.

23. The computer-readable storage medium of claim 22, which when executed cause the system to further compare the first measure with the second measure to determine an impact of the one or more events with respect to the issue.

24. The computer-readable storage medium of claim 22, wherein producing the first and second measures regarding the cases associated with the issue comprises producing the first and second measures regarding the cases associated with a problem, and

wherein receiving the information after the one or more events have occurred comprises receiving the information after a fix has been made available for the problem.

25. The computer-readable storage medium of claim 22, wherein producing the first and second measures is performed by a quantifier associated with the trained categorizer.

26. The computer-readable storage medium of claim 22, wherein producing the first measure comprises:

computing an actual measure; and

adjusting, based on at least one factor, the actual measure to produce the first measure.

27. The computer-readable storage medium of claim 26, wherein adjusting the actual measure comprises extrapolating the measure for a future time period.

28. The computer-readable storage medium of claim 22, wherein the one or more events comprise one or more fixes for the issue, the instructions which when executed cause the system to further:

compare the first and second measures;

receive cost information relating to the one or more fixes; and

report a return on investment based on comparing the first and second measures and the cost information.

35. The computer-readable storage medium of claim 22, wherein developing the training cases comprises:

adding at least one of the identified cases to a positive training set in response to receiving user confirmation; and

adding at least one of the identified cases to a negative training set in response to receiving user disconfirmation,

wherein the developed training cases include the positive training set and the negative training set.

29. A method comprising:

computing, based on output from a categorizer executed on one or more processors, a first measure regarding a first set of cases associated with a first issue and a second measure regarding a second set of cases associated with the first issue, wherein the first set of cases is associated with a first time period prior to occurrence of an event, and wherein the second set of cases is associated with a second time period after occurrence of the event;

comparing, by the one or more processors, the first and second measures to determine an impact of the event on the first issue;

identifying plural issues including the first issue;

selecting the first issue from among the plural issues based on an expected return on investment; and

effecting action with respect the first issue, wherein the event comprises the action.

30. The method of claim 29, wherein determining the impact of the event on the issue comprises determining the impact of an action with respect to a customer support issue.

31. A system, comprising:

one or more processors;

a categorizer executable on the one or more processors;

a search engine to identify at least one case in response to a query relating to an issue;

an interface to receive an indication that the identified at least one case belongs to the issue; and

a training module executable on the one or more processors to train the categorizer based on the received indication;

the trained categorizer executable on the one or more processors to classify cases in a data set into the issue;

a quantifier executable on the one or more processors in cooperation with the trained categorizer to compute a first quantification measure for a first time window, and a second quantification measure for a second time window, wherein the first time window is prior to occurrence of an event corresponding to the issue, and the second time window is after occurrence of the event; and

a module executable on the one or more processors to determine an impact of the event on the issue based on comparing the first and second quantification measures.

36. The system of claim 31, wherein the training module is to train the categorizer using a positive training set of cases and a negative training set of cases, wherein the training module is to add the identified one case to the positive training set in response to receiving the indication that the identified at least one case belongs to the issue, and wherein the training module is to add the identified one case to the negative training set in response to receiving the indication that the identified at least on e case does not belong to the issue.