A. Mousavi, B.H.Far (Canada), and A. Eberlein (UEA)
Scenario-based specifications, domain theory, behaviour
modeling, emergent behaviours.
A new approach for generating state machine designs from
scenarios is presented that assigns state values to the states
of system’s components (processes) in different scenarios.
State values are assigned using a light domain theory in
ferred from the domain knowledge, and are used to detect
identical states of processes in order to merge partial be
haviours of scenarios.
The domain theory will be systematically constructed
by requesting the domain expert to look at some tables that
their rows and columns are filled with selected messages
from scenarios, and possibly finds one cell that its column
has a special relation with its row called semantical causal
ity. Semantical causality captures an invariant property of
a system in terms of performance dependability between
messages and as a part of the domain knowledge that is not
explicitly defined in a scenario.
Furthermore, to detect emergent behaviours in the
generated state machines, non-deterministic behaviour of
processes is defined to charactrize the conditions that emer
gent behaviours are allowed by systems’s architecture de
fined by scenarios.