| 摘要 |
Failure messages generated as a result of tests performed on a target application are bucketed in an effort to correlate related failure messages with minimal or no human interaction. An exemplary method for bucketing failure messages includes receiving a failure message, searching a stack trace of a target application for a blame stack trace frame from which the failure message originated, finding a most similar match bucket from a plurality of buckets, determining if the most similar match bucket at least meets a similarity threshold, adding the failure message to the most similar match bucket, if it is determined that the most similar match bucket at least meets the similarity threshold, and creating a new bucket and adding the failure message to the new bucket, if it is determined that the most similar match bucket does not at least meet the similarity threshold.
|
