Facilitating process analysis through visualising process history: experiences with a Dutch municipality

Nowadays vast quantities of data are stored as a result of the operation of software systems and devices. The analysis of this data can provide valuable insights. In the field of Business Process Management, event logs may provide valuable information for business process improvement. This is the realm of process mining, an area which has provided many analysis techniques over the past decade. Despite the abundance of process mining techniques, it remains a challenge to provide results that are understandable by domain experts. Discovered process models are often perceived as abstract and static. Conformance checking techniques provide detailed results that are only understandable for process analysts. Therefore, we propose an approach to dynamically visualize event data on intuitive 'maps'. States of the process are visualized on a collection of maps thus resulting in sequences of 'photographs' of the process under investigation. By replaying the event log using such visualizations we can create a collection of 'process movies'. Our visualisation approach has been implemented in ProM and allows for any type of 'map' as long as activity instances can be associated to map coordinates. Moreover, the approach has been evaluated in collaboration with a Dutch municipality