In the present document we present the first part of the DiVA RE approach which mainly focuses on answering the question on what (requirements, their justification and context) to model, as well as touches on the issue of how to restrict the proliferation of variants in requirements models. To address these questions, the DiVA RE approach proposes to:

• construct a feature tree from the input requirements text, thus identifying specific requirements that need to be modelled, and refining this tree with variations and adaptation points;

• recover the hard- and soft-goals related to the given requirements to provide a justification for them, and also to use a single goal-model for modelling the large set of potential variations;

• analyse the resultant goal/feature tree to extract the relevant context and related variability and constraints which restricts the set of potential configurations.

Open Document