Independently of the specific task to be enacted in a smart space, it is always crucial to mine a set of models representing environmental dynamics and, noteworthy, user habits and desires. Many different formalisms have been proposed to model human habits, but the vast majority of them are either difficult to read and evaluate or their definition requires a huge amount of work from either experts or users. In this paper we propose to employ process mining techniques in order to model human habits, and we experimentally evaluate such an approach on a dataset built adopting the Smart-Home-in-a-Box toolkit with real users.