Changes and evolution of software systems constantly gener- ate new challenges for the recovery of software systems architectures. A system's architecture, together with its elements and the way they inter- act, constitute valuable assets for understanding the system. We believe that offering multiple architectural views of a given system, using domain and pattern knowledge enhance understanding of the software system as a whole. To correlate different sources of information and existing soft- ware system, different viewpoints are considered. Viewpoints enable one to model such information and guide the extraction algorithms to ex- tract multiple architectural views. We propose a recursive framework, an approach that expresses different kinds of information as viewpoints to guide the extraction process. These multiple viewpoints models al- low considering architectural, conceptual, and structural aspects of the system.