Mining configurable enterprise information systems

Process mining is the extraction of a process model from system logs. These logs have to meet minimum requirements, i.e. each event should refer to a case and a task. Many system logs do not meet these requirements, and therefore it is not possible to use process mining for process optimization or delta analysis. This paper shows an alternative process mining procedure for logs containing data on the frequency that process steps have been executed. To be able to mine such logs we apply Configurable Event-driven Process Chains (C-EPCs). If a C-EPC is available, we propose a method to mine the process. If only a classical reference model (i.e. an EPC) is available, we propose a method to first derive the C-EPC through mining and then analyse the process. This approach enables us to do process mining in the context of ERP systems such as SAP R/3.