Process mining : overview and outlook of Petri net discovery algorithms

Within the research domain of process mining, process discovery aims at constructing a process model as an abstract representation of an event log. The goal is to build a model (e.g., a Petri net) that provides insight into the behavior captured in the log. The theory of regions can be used to transform a state-based model or a set of words into a Petri net that exactly mimics the behavior given as input. Recently several papers appeared on the application of the theory of regions for process discovery. This paper provides an overview of different Petri net based discovery algorithms from both the area of process mining and the theory of regions. The overview encompasses five categories of algorithms, for which common assumptions and problems are indicated. Furthermore, based on the shortcomings of the algorithms in each category, a set of directions for future research in the process discovery area is discussed.