Mapping Graph-Based Process Model Into Discrete Event Simulation (DES)

Many process discovery algorithms are emerging to form process models based on event logs automatically. There are several representations of the process model, one of which is the graph-based process model. The advantage of using a graph-based process model is that event logs, and process models can be represented in a graph model, so there is no conversion cost to converting event logs to process models. However, the weakness of the graph-based process model is that there are no features to simulate the model. To the contrary, Discrete Event Simulation (DES) is a type of process model that can be used for simulation; However, DES is created manually based on the knowledge of experts. This study proposes mapping a graph-based process model to the DES model. This research also makes an algorithm to determine the time span of the implementation of activities, which is used in the DES model. This study generates event logs based on DES obtained using the proposed mapping and builds a graph-based process model based on event logs. The evaluation verifies that the graph-based process model is the same as the existing one, which means that the proposed mapping is successful.