Workflow mining: a survey of issues and approaches
Data & Knowledge Engineering
Workflow Mining: Discovering Process Models from Event Logs
IEEE Transactions on Knowledge and Data Engineering
Genetic process mining: an experimental evaluation
Data Mining and Knowledge Discovery
Conformance checking of processes based on monitoring real behavior
Information Systems
Modularity in Process Models: Review and Effects
BPM '08 Proceedings of the 6th International Conference on Business Process Management
Recovering business processes from business applications
Journal of Software Maintenance and Evolution: Research and Practice
A discourse on complexity of process models
BPM'06 Proceedings of the 2006 international conference on Business Process Management Workshops
A fast and elitist multiobjective genetic algorithm: NSGA-II
IEEE Transactions on Evolutionary Computation
Domain-driven reduction optimization of recovered business processes
SSBSE'12 Proceedings of the 4th international conference on Search Based Software Engineering
| Hi-index | 0.00 |
While models are recognized to be crucial for business process management, often no model is available at all or available models are not aligned with the actual process implementation. In these contexts, an appealing possibility is recovering the process model from the existing system. Several process recovery techniques have been proposed in the literature. However, the recovered processes are often complex, intricate and thus difficult to understand for business analysts. In this paper, we propose a process reduction technique based on multi-objective optimization, which at the same time minimizes the process complexity and its non-conformances. This allows us to improve the process model understandability, while preserving its completeness with respect to the core business properties of the domain. We conducted a case study based on a real-life e-commerce system. Results indicate that by balancing complexity and conformance our technique produces understandable and meaningful reduced process models.