Mining Resource Community and Resource Role Network From Event Logs

Abstract

This paper deals with the perspective of organization based on process mining, which concentrates on activities and performers (resources) in event logs. A classic social network only reflects how closely related any two resource individuals are. Some resources may play similar roles in the network, and they can establish their own resource community. The algorithm reported in this paper constructs a resource community net with optimal modularity. Moreover, we find that there are some resources playing multiple roles in different communities, taking on different responsibilities in a system. Another algorithm is developed to discover these multi-role resources, and a corresponding resource role network is built. Finally, two examples are presented to demonstrate the proposed approach.