A Meta Model for Software Architecture Conformance and Quality Assessment

Abstract

Software architecture and design suffer from a lack of documented knowledge on how different architectural styles influence software quality. Existing software quality models do not allow engineers to evaluate whether a given software system adequately implements the basic principles of the chosen architectural style, and which architectural properties and best practices beyond these principles contribute to the system's quality. In this paper, I present a meta quality model for software architectures, which can be used not only as a knowledge-base to easily compare architectural styles based on their impact on software quality, but also to increase efficiency of architectural quality analysis by leveraging existing modeling concepts and tools. An experiment performing an architecture assessment using a quality model for the SOA architectural style not only showed that the approach is applicable in practice, but also indicated a reduction of manual effort compared to other architecture assessment approaches.