This dissertation investigates algorithmic techniques for temporal process discovery in many domains. Many different formalisms have been proposed for modeling temporal processes such as motifs, dynamic Bayesian networks and partial orders, but the direct inference of such models from data has...