Methods, systems, and computer program product for automatically categorizing defects

Patent metadata

Field	Value
Publication number	US-9020943-B2
Application number	US-35021709-A
Country	US
Kind code	B2
Filing date	Jan 7, 2009
Priority date	Jan 7, 2009
Publication date	Apr 28, 2015
Grant date	Apr 28, 2015

Abstract

Disclosed are method(s), system(s), and computer program product(s) for automatically categorizing a defect into a category based at least on a keyword related to the defect. The method identifies information regarding the defect and identifies additional information related to categorizing the defect. The method identifies a keyword and determines whether the defect may be categorized into a category using the keyword. The method further comprises determining if the result of categorization is deterministic. In some embodiments, the method or the system further comprises associating a first score for categorizing the defect into the category based on the keyword and a second score for categorizing the defect into the category based on another keyword. In some embodiments, the method or the system further comprises identifying or determining a relationship between the first score and the second score with respect to the first category for the defect.

First claim

We claim:

1. A machine-implemented method for automatically categorizing software product defects into two or more categories, the method comprising:
identifying first information regarding a first defect;
identifying or receiving additional information from a data structure, wherein the additional information is related to an act of categorizing the first defect into a first category, wherein the data structure comprises a rule base which comprises at least one inference rule to mimic a reasoning process of a human expert, and wherein the data structure comprises a keyword database;
identifying a first keyword for the first defect based at least in part upon the first information;
identifying a plurality of weights for the first keyword wherein values for the plurality of weights are specific to a combination of a given keyword and a given category;
accessing, from the keyword database, a first weight for the first keyword with respect to a first category;
accessing from the keyword database, a second weight for the first keyword with respect to a second category;
categorizing the first defect into the first category, based at least in part on the first weight for the first keyword with respect to the first category; and
categorizing the first defect into the second category based at least in part on the second weight for the first keyword with respect to the second category;
wherein the second category is distinct different from the first category;
wherein the first weight is distinct different from the second weight; and
updating the data structure with a result of an act of categorizing the first defect into the first category.

2. The machine-implemented method of claim 1 , further comprising:
categorizing the first defect into a third category, the third category being distinct from the first category.

3. The machine-implemented method of claim 1 , wherein the data structure comprises at least one of, a defect database, a defect detail database, a category database, a category-defect database, or a defect tracking database.

4. The machine-implemented method of claim 1 , wherein the additional information from the data structure comprises metadata for the first defect, diagnostic data for the first defect, a description of one or more symptoms of the first defect, an error message caused by the first defect, one or more snippets of a log file or a trace file, an engineer comment about the first defect, or a customer response to or description of the first defect.

5. The machine-implemented method of claim 1 , wherein the first information regarding the first defect comprises a service request, a defect report, or a defect analysis report.

6. The machine-implemented method of claim 1 , further comprising:
determining if the first defect may be categorized into the first category by further using a second keyword.

7. The machine-implemented method of claim 6 , further comprising:
associating a first score for categorizing the first defect into the first category based at least in part on the first keyword and a second score for categorizing the first defect into the first category based at least in part on the second keyword.

8. The machine-implemented method of claim 7 , further comprising:
identifying or determining a function describing a relationship between the first score and the second score with respect to the first category for the first defect.

9. The machine-implemented method of claim 8 , wherein the function describes the relationship between the first score and the second score by performing a weighted sum for the first score and the second score.

10. The machine-implemented method of claim 8 , wherein the function describes a non-linear relationship between the first score and the second score.

11. The machine-implemented method of claim 1 , further comprising:
expanding the rule base, wherein a desired result of categorization is not produced.

12. The machine-implemented method of claim 11 , wherein the act of expanding the rule base is performed without modifying other programs for categorization.

13. The machine-implemented method of claim 1 , further comprising:
a problem solving program for categorization which comprises expertise or expert knowledge of the first defect.

14. The machine-implemented method of claim 13 , further comprising:
modifying the problem solving program, wherein a desired result of categorization is not produced.

15. A computer program product comprising a non-transitory computer-usable medium having executable code to execute a process for automatically categorizing software product defects into two or more categories, the process comprising:
identifying first information regarding a first defect;
identifying or receiving addition information from a data structure, wherein the additional information is related to an act of categorizing the first defect into a first category; wherein the data structure comprises a rule base which comprises at least one inference rule to mimic a reasoning process of a human expert, and wherein the data structure comprises a keyword database;
identifying a first keyword for the first defect based at least in part upon the first information;
identifying a plurality of weights for the first keyword wherein values for the plurality of weights are specific to a combination of a given keyword and a given category;
accessing, from the keyword database, a first weight for the first keyword with respect to a first category;
accessing from the keyword database, a second weight for the first keyword with respect to a second category;
categorizing the first defect into the first category, based at least in part on the first weight for the first keyword with respect to the first category; and
categorizing the first defect into the second category, based at least in part on the second weight for the first keyword with respect to the second category;
wherein the second category is distinct different from the first category; and
wherein the first weight is distinct different from the second weight.

16. A system for automatically categorizing software product defects into two or more categories, the system comprising:
a communication module configured for identifying first information regarding a first defect;
a data interface configured for identifying or receiving addition information from a data structure, wherein the additional information is related to an act of categorizing the first defect into a first category, wherein the data structure comprises a rule base which comprises at least one inference rule to mimic a reasoning process of a human expert, wherein the data structure comprises a keyword database;
a processor configured for identifying a first keyword for the first defect based at least in part upon the first information, and identifying or receiving a plurality of weights for the first keyword wherein values for the plurality of weights are specific to a combination of a given keyword and a given category;
the processor or a second processor configured for accessing, from the keyword database, a first weight for the first keyword with respect to a first category;
accessing from the keyword database, a second weight for the first keyword with respect to a second category;
the processor or a third processor configured for categorizing the first defect into the first category, based at least in part on the first weight for the first keyword with respect to the first category; and
the processor or a fourth processor configured for categorizing the first defect into the second c

Assignees

• Caldwell Bob
• Rao Sandesh
• Burton William
• Ort Joshua
• Brewer-Tillotson Mark
• Oracle Int Corp

Inventors

• Caldwell Bob
• Rao Sandesh
• Burton William
• Ort Joshua
• Brewer-Tillotson Mark

Classifications

• G06F16/285 Primary

  Clustering or classification · CPC title

• G06F17/30598 Primary

  Physics · mapped topic