Software quality properties' normally derived from non-functional requirements' are becoming more important for software. A main reason for software evolution is the unsatisfaction to software quality properties. When improving these properties through software evolution' it is essential to know whether software functions are affected and by how much. This paper proposes an approach to linking the functions with the quality properties of software for evolution via software architecture styles' aiming at contributing to (1) predicting evolution efforts and (2) transforming software for improving its quality. ? 2012 IEEE.
