Software evolution control towards a better identification of change impact propagation

The software evolution is often a continuous process necessary to avoid a short longevity of software use. The evolution control requires analysis to be processed respectively before and after software change implementation. First, an a priori change impact analysis is required to help decisions takers by estimation of intended impact and its propagation to other software components. Secondly, an a posteriori change analysis is destined to estimate to what extent the goals of implemented changes have been effectively reached. Both invoked analysis of change impact require an exhaustive description of individual software components as well as their various types of interdependency. In our paper, we propose an exhaustive stratified description of software components using formal modeling, aiming at providing better change impact analysis. The proposed description focuses on software elements which may play role in change impact and the propagation of its effects from directly modified component to other components of the whole software.