An applicable methodology to automate software quality measurements

Abstract

Numerous software metrics have been presented in the recent years. However, even if they seem to be able to provide many advantages in the improvement of software quality, they have not helped in terminating the so called "software crisis" . The reason for this is not the inability of metrics to improve the software quality, but the ineffective use of metrics and the lack of easy-to-use and automated methods to collect, interpret and apply the measurement results. We present a methodology of applying three software metrics in a completely automated way, so as to collect and interpret the output measurements and to use the conclusions, in order to improve the quality of the measured software. The experimental use of this methodology, in separate life cycle phases of two sample software projects, is also presented.