The Research Of Software Defect Metrics And Implementation Of Defect Inspection System Based On SPC

Process control and management is a crucial tache to ensure the product quality. Statistical process control technology, whose applications has exceeded the industrial field, is a scientific management method of ensuring preventive principle as well as carrying out the process control. How to measure the software process metrics has already been one of the hot topics on computer software engineering research. So the software process metrics based on statistical process control technology is a good question worthy of probing into.The target of this paper is to apply the statistical process control technology tothe software process metrics. Firstly, it begins with in-depth research on the softwareprocess metrics theories to lay a theoretical foundation for the software defect metrics.The Capacity Maturity Module (CMM) is also exploited to bring forward a referencestandard for it. Secondly, some diversified key technologies connected with the objectof defect are studied, including the definition of defect, attribute settings, metricsmeasurements and methods. Meanwhile, based on investigating the statistical processcontrol technology deeply and completely, it mainly researched into the control chart,judging criterions and the definition of process abilities or computing methods, andproposes a novel defect metrics method based on Shewhart control chart. Byanalyzing instances, this paper also exploits the applications of X — R control chart,X—S control chart and U control chart in software defect metrics. Finally, byintegrating the outcomes of those researches mentioned above, the software defectinspection system has been implemented, which is able to track and show the defects,defect data update,data neaten,analysis stat.,compile and stat., as well as create reportforms and control charts.The results of our experiments show that the metrics of software defect has a certain effect on ensuring the software process quality, and applying the statistics process control technology to the defect metrics is feasible and effective. As to improving software process and software qualities, what this paper has done deals with only one aspect. Thus, there are still a lot of things to be researched in the future, and the application of the statistics process control technology should be enlarged to ensure software process quality more effectively.