The analyst’s job can be seen as mapping the real world to a model. At this an analyst has two main ways of making use of Model-to-Model Transformation in requirements analysis:
1. Applying rules to models
Looks like the state machine in figure 1 has some, say, medium level of complexity. But when broken down to model + rules it becomes much simpler. Where the model is the one in figure 2 and the rule says “every state transition between these processing state needs to be 4-eyes-checked”.
Thus one gains the understanding that 1, 3, 5, 7 are processing states (where sth is being processed) and 2, 4, 6 are checking states (where some verification is done). Even if one has exceptions to the rule like the edge 5-3, where no check is performed, the triple model + rule + exception still provides a much better understanding than the model alone.
So, all uses of Model-to-Model Transformation in Analysis should belong to one of the above categories – do they?