Workflow mining: Discovering process patterns & data analysis from MXML logs

Abstract

Contemporary workflow management systems are determined by precise process models, i.e., a totally specific workflow design is necessary so as to discover a given workflow process. Creating a workflow design is a complex time-consuming process and characteristically there are conflicts and discrepancies between the real-time workflow processes and the processes as professed by the management. Consequently, we recommend a method for process mining. This method employs workflow logs to determine the workflow process as it is truly being implemented. The process mining method proposed in this paper can cope with noise and can also be exercised to authorize workflow processes by revealing and measuring the discrepancies between prescriptive models and factual process executions.