Finding Structure in Unstructured Processes: The Case for Process Mining

Today there are many process mining techniques that allow for the automatic construction of process models based on event logs. Unlike synthesis techniques (e.g., based on regions), process mining aims at the discovery of models (e.g., Petri nets) from incomplete information (i.e., only example behavior is given). The more mature process mining techniques perform well on structured processes. However, most of the existing techniques fail miserably when confronted with unstructured processes. This paper attempts to "bring structure to the unstructured" by using an integrated combination of abstraction and clustering techniques. The ultimate goal is to present process models that are understandable by analysts and that lead to improved system/process redesigns.