Error Metrics for Business Process Models
Little research has been conducted so far on causes for errors in Business Process Models. In this paper, the authors investigate on how mainly domain independent factors such as the size or complexity of models influence errors observed in a wide range of existing Business Process Models. In particular, they provide a set of six metrics presumably related to the comprehensibility of both the process model structure and the process state space, and discuss their capability to predict errors in the SAP reference model. The results show that already the three metrics size, separability, and structuredness suffice to achieve a high Nagelkerke R2 value of 0.853 demonstrating a good predictive efficacy.