Language-Based Process Model Discovery and Enhancement

Introducing process models into an organization entail many advantages even more when being adapted to the organizations specific environment and needs. Unnecessary processes may be omitted and never be planed or defined for execution, while the necessary ones consist of possible workflows in an organization. Nevertheless, lived processes deviate from their correlated process definitions due to evolution in process execution. Thus, an aggregation of lived processes, forming a process model describing them, is needed for comparison with the organizational process model. Differences between these models provide useful information for process improvement. Therefore, a language-based approach is used to construct a process model as deterministic finite automaton describing the executed processes. An evaluation is done on real workflow data of a German manufacturer. It results in difficulties and restrictions of the proposed approach, which will be discussed.