Graph-based Process Mining for Measuring Quality of Business Process Model

Process mining can be used to model business processes. In process mining, graph-based can be used as an option in making a process model that is efficient based on quality. This research proposes a method of measuring fitness, precision, simplicity, and generalization. Measurement evaluation is carried out using two processes, port container handling and waste handling. Based on the research, the results of the fitness and precision port container handling using the Proposed Graph-based Quality Measurement and ETM Quality Measurement methods are 1.00. Meanwhile, the generalization obtained from both methods is 0.96. Simplicity generated by the Proposed Graph-based Quality Measurement method is 0.87 meanwhile the result of ETM Quality Measurement method is 1.00. In the case study of waste handling, based on both methods, the fitness is 1.00 and the generalization is 0.83. Precision obtained from the Proposed Graph-based Quality Measurement method is 0.95 and 0.75 from ETM Quality Measurement method. The simplicity obtained from the Graph-based Quality Measurement method is 0.95 meanwhile from ETM Quality Measurement method is 1.00. The simplicity calculation results are different because ETM cannot detect invisible tasks, while Graph-based Quality Measurement can discover them.