Collaborative information systems are becoming more and more complex, involving numerous interacting business objects within considerable processes. Analysing the interaction structure of those complex systems will enable them to be well understood and controlled. The work described in this paper is a contribution to these problems for workflow based process applications. In fact, we discover workflow patterns from traces of workflow events based on a workflow mining technique. Workflow mining proposes techniques to acquire a workflow model from a workflow log. Mining of workflow patterns is done by a statistical analysis of log-based event. Our approach is characterised by a “local” workflow patterns discovery that allows to cover partial results and a dynamic technique dealing with concurrency.