Research on a Workflow Mining Algorithm Using Dynamic Weighted Association Rules Based on Tuple of Activity-Role

In order to obtain workflow model made of tuples of activity-role from work logs, this paper presents a workflow mining algorithm. Firstly, identify the frequent sets with nodes based on tuples of activity-role, through the weighted association rules algorithm which avoid missing of some of the important but non frequent role through setting the different values of the roles. The node flow matrix can be lead from merged frequent nodes sets. The workflow model which consists of tuples including the activities and roles are practical and safe full-featured.