Extended Abstract

Since 1996, various attempts have been made to unify different object-oriented modeling languages. As a result of this endeavor, two languages have been developed: the Unified Modeling Language UML [3] and the Open Modeling Language OML [2]. Both UML and OML introduce a distinctive new feature: they allow users to extend or even to modify the base language in order to adapt the language to specific situations or needs. The language construct that is used to implement this feature is called a stereotype . In the context of object-oriented modeling, the notion of stereotypes was introduced before by Rebecca Wirfs-Brock [4]. Her principal idea is to provide a secondary classification for objects: stereotypes classify objects according to their use, independently of the primary classification by classes and class inheritance. UML and OML both generalize WirfsBrock's notion of stereotypes from a secondary classification to a concept that allows for general extensions of the base language. A stereotype in UML and OML can add new properties to elements of the underlying language or can modify existing ones. In our contribution we discuss the UML/OML kind of stereotypes in a general context of object-oriented modeling languages. However, as the notion of stereotypes is not limited to object-oriented approaches, we define a stereotype independently of object-orientation as follows.