This paper deals with reusing of software process models. Based on the insufficiencies of existing software process reusing approaches (limited reusability of the software process components), we propose a new approach that promotes a large reuse of existing proven software process models even not oriented components. Our approach is based on two steps: we use domain ontology to capitalize the software process knowledge and we handle the inferred knowledge as software process architecture. In this paper, we present a related works on this field and introduce the general outlines of our approach that is under validation.