Requirements Scenarios Formalization Technique N Versions Towards One Good Version

Using a notion of simulation introduced by de Roever et al. [9], we define a corresponding notion of bisimulation which we will use as a test to detect formality discrepancy between two formal versions of the same informal scenario. We will also use it, combined with the relation part of, as a test to detect discrepancy between two formal scenarios describing two different informal scenarios. We will describe a way of implementing these formality discrepancy tests in the work place. We call it N-version technique for scenario formalization. This technique allows us to (1) detect formality discrepancy between formal scenario versions corresponding to the same scenario, to (2) validate, for the given scenario, the formalization obtained by a specifier against another obtained by another specifier before bothering the stake-holder with problems of the specifier's creation (formalization errors), and to (3) detect formality discrepancy between the formal scenarios of different informal scenarios if their respective spaces share at least one component.