The measurement of software science parameters in software designs

Abstract

Metrics of software quality have historically focused on code quality despite the importance of early and continuous quality evaluation in a software development effort. While software science metrics have been used to measure the psychological complexity of computer programs as well as other quality related aspects of algorithm construction, techniques to measure software design quality have not been adequately addressed. In this paper, software design quality is emphasized. A general formalism for expressing software designs is presented, and a technique for identifying and counting software science parameters in design media is proposed.