Identifying bottlenecks and fraud of business process using alpha ++ and heuristic miner algorithms (Case study: CV. Wicaksana Artha)

摘要

Every organization or company has a business process to support the structured running tasks. Each of the business processes has a role called resource that is useful for running the processes. However, a few organizations and companies evaluate their business processes. Process mining is a way to evaluate event log data by forming a workflow. Petri net is chosen workflow to analyze the occurrence of bottlenecks. A bottleneck is an event that has a longer waiting time than the other business processes. PROM is used to analyze the business process in mxml formal by using process mining algorithm, such as alpha++ and heuristic miner algorithms. Both algorithms are used to determine the difference of occurring bottlenecks. This research shows that the algorithm affects the location of the occurrence of bottleneck. The location of the bottleneck is seen based on the calculation of the token time in the place (place between 2 transitions or transactions). Heuristic miner can model a business process appropriately and it has a higher performance than alpha++ algorithm because it can calculate the frequency, whereas alpha++ algorithm can not generate the frequency calculation. The heuristic miner algorithm can also check the fraud by utilizing its fitness value. The obtained fitness value will indicate whether there are indications of fraud or no indication in every business process.