Unifying Multi-level Business Process Discovered by Heuristic Miner Algorithm

Process mining techniques are commonly used in business management while conformance checking becomes an important issue in business process management. Researchers derive the actual business process from event logs to draw a comparison with the business process model. When the two are inconsistent, a lack of internal controls happens. This research proposes the consistence checking in the event logs. Because of the different granularity in the same event logs, the process can be demonstrated as grain 1 to grain n, in which the smaller the grain means the finer the granularity of the process. While using a process log to retrace the business process, different business processes might be shown in the processes with different granularities in the same event logs. The dependency threshold of Heuristic miner algorithm is used to deal with the differences of consistency automatically in this research. This research uses the event logs from a marble processing industry for the case conformance. Focusing on the fine and coarse two granularities of the business process matrix, the conformance checking is applied for a consistent business process via the setting of the dependency threshold and the consistent ratio. The result show the valuable information to audit or re-design the business process model.