Optimized generation of data for software problem analysis

Abstract

A computer optimizes the prospective generation of data used for analysis of a software problem. The computer generates data in accordance with data generation parameters and a software problem is analyzed with reference to the data so generated. The problem analysis produces a report that details specifics of the software problem, the data that was available for analysis, a flag to indicate success or failure of the analysis to identify a root cause, and information about whether the data supplied was insufficient, sufficient, or superfluous with respect to identifying a root cause of the software problem. The method then uses the analysis report to modify the data generation parameters, thereby iteratively optimizing the data that are generated for analysis of subsequent software problems.

What is claimed is: 
     
         1 . A method for automatically optimizing prospective generation of data used for analysis of a software problem the method comprising:
 generating, by one or more processors, data for analysis of a software problem, in accordance with data generation parameters;   analyzing, by one or more processors, a software problem, based at least in part, on data generated for analysis of the software problem, using one or more computer-implemented methods including cognitive computing, image analysis, character recognition, and pattern recognition to produce at least an analysis report configured to provide one or more of:
 a description of the software problem; 
 a Boolean flag indicating success or failure of the analysis to identify a root cause of the software problem; 
 a description of data that was supplied for the analysis, including data types and respective quantities thereof; 
 a description of data types, or respective quantities thereof, provided for the analysis, but not needed for a successful analysis; and 
 a description of data types, or respective quantities thereof, needed for a successful analysis, but not provided; 
 saving the analysis report in an analysis history; 
 modifying, by one or more processors, the data generation parameters in accordance with one or more rules, to create modified data generation parameters; and 
 generating, by one or more processors, data for analysis of the software problem, in accordance with the modified data generation parameters. 
   
     
     
         2 . The method of  claim 1 , further comprising:
 iteratively optimizing, by one or more processors, data generation parameters, by utilizing a feedback loop, to enable software problem analysis to converge on a solution of the software problem.   
     
     
         3 . The method of  claim 1  wherein modifying data generation parameters comprises:
 evaluating, by one or more processors, an optimizing rule based at least in part on one or more of the analysis report, or the analysis history and wherein the optimizing rule can be modified in response to evaluation of an optimizing rule, problem determination results, a new or updated optimization goal, or other aspects of a computer system. 
 
     
     
         4 . The method of  claim 3  wherein the analysis history comprises:
 an accumulation of one or more analysis reports pertaining to one or more software problem analyses. 
 
     
     
         5 . The method of  claim 3  wherein the optimizing rule comprises:
 a Boolean conditional; and 
 one or more actions. 
 
     
     
         6 . The method of  claim 3  wherein evaluating an optimizing rule further comprises:
 evaluating, by one or more processors, a Boolean conditional of an optimizing rule; and 
 performing, by one or more processors, actions of the optimizing rule if the Boolean conditional evaluates to a pre-determined state.