Discovering Process Reference Models from Process Variants Using Clustering Techniques

In today's dynamic business world, success of an enterprise increasingly depends on its ability to react to changes in a quick and flexible way. In response to this need, process-aware information systems (PAIS) emerged, which support the modeling, orchestration and monitoring of business processes and services respectively. Recently, a new generation of flexible PAIS was introduced, which additionally allows for dynamic process and service changes. This, in turn, has led to large number of process and service variants derived from the same model, but differs in structures due to the applied changes. This paper provides a sophisticated approach which fosters learning from past process changes and allows for determining such process variants. As a result we obtain a generic process model for which the average distances between this model and the process variants becomes minimal. By adopting this generic process model in the PAIS, need for future process configuration and adaptation will decrease. The mining method proposed has been implemented in a powerful proof-of-concept prototype and further validated by a comparison between other process mining algorithms.