Some remarks on a class of inverse problems related to the parabolic approximation to the Maxwell equations: a controllability approach