From the i* diversity to a common interoperability framework

Abstract

Since the i* framework emerged in the requirements engineering community in the mid-90s, different research groups have proposed variations to the modelling language proposed in the i* framework (¿the i* language¿). There are basically two reasons behind this fact: ¿ The definition of the i* language is loose in some parts, and some groups have opted by different solutions or proposed slight changes to the original definition. ¿ Some groups have used the i* framework with very different purposes, thus different concepts have become necessary, many of them have added their own modelling constructs. As a consequence, several i* variations at different maturity levels have been produced.

The adaptability of i* to different needs is part of its own nature, therefore these variations are not to be considered pernicious, on the contrary, flexibility may be considered one of the framework¿s key success features. In this context, the objective of this thesis is to tackle the i* Interoperability Problem: to propose a framework to understand the variations of the i* modelling language and, considering this framework, to generate a proposal to support the interoperability and integration of these variations.

To do so, the PhD thesis has identified and solved up to ten research questions related to the following categories: 1) The i* variants have been identified and characterized. Metamodels have been used to align these variants and identify an appropriate kernel. A critical look on the ¿why¿ this happened has been conducted from several perspectives. 2) An interoperability framework in terms of metamodel alignment has been formulated. This interoperability framework has come from the adaptation of a solution to metamodel evolution and model co-adaptation formulated in the domain of model-driven engineering. 3) An XML-based interoperability format implementing the framework, iStarML, has been deployed. 4) The solution has been validated in terms of community acceptance by means of an empirical study, and from a practical point of view by the interconnection of different existing tools. The set of related publications shows the research path, starting from the identification and characterization of the interoperability problem, pass trhought a theoretical explanation that sustaints a technical proposal, and ends illustrating inteoperation cases and its community acceptation.