Towards a Better Change Impact Analysis in Architecture Description Languages

Abstract

This chapter proposes a multi-modeling approach destined to better control the software evolution. The presented approach follows formal models on software architecture and source code level. It formalizes the elements of software architecture, their interdependent relationships and their source codes to analyze the impact propagation of an intended change. The constituents of these models are evaluated with a reasoning based expert system. The expert system is validated as a platform based on eclipse plug-ins to analyze the architecture description languages. The software architecture and source codes are parsed to generate the facts of the distributed knowledge-based system, which executes change propagation rules to evaluate the impact of a change performed on distributed components. © Springer-Verlag Berlin Heidelberg 2011.