A workflow mining tool based on logs statistical analysis

As more and more Workflow management systems (WfMS)s log every event occurring during the process execution, customers become increasingly interested in managing process executions Specifically, there is a desire for getting more visibility into process executions to be able to quickly spot problems Our contribution is an original statistical technique to discover workflow patterns from event-based logs Indeed, our approach is characterised by a "local" workflow patterns discovery that allows to cover partial results through a dynamic programming algorithm Furthermore, we improve the discovery of the activities concurrent behaviour by proposing a dynamic mechanism of logs window sliding Our approach has been implemented within the ProM framework, as a plug-in.