Making explicit some implicit i* language decisions

Abstract

The i* (i-star) framework is one of the most widely adopted modelling approaches by several communities (business modelling, requirements engineering, ...). Probably due to its highly strategic nature, the definition of the modelling language offered by the framework does not make explicit the full behaviour of some basic constructs, leaving them thus open to several interpretations. This looseness may not be important in some contexts, even it may be beneficial since it leaves room for researchers to customize the framework to their needs. However, it becomes an obstacle in other situations, e.g., model interoperability and model-driven development. In this paper we identify ambiguities and silences in the i* language definition in a systematic manner, and then we propose an interpretation to deal with them. In some cases, the proposal may include the addition of some annotation into some language construct. The result is a formal definition taking the form of a UML conceptual data diagram (a metamodel) with several important integrity constraints. © 2011 Springer-Verlag.