Bayesian Network for finding best combination of performers in BPM environment

Business process management system (BPMS) is an essential element in the business environments as a technology for delivering services to customers. In order to satisfy the customers, Business Process (BP), as a series or network of value-added activities, should be performed by relevant performers. One of the mechanisms to manage performers in BPMS is Separation-of-Duty (SoD) authorization policy under role-based access control (RBAC) mechanism. In the case of dynamic SoD, user assignment based on the constraints occurs in the run-time environment that results various execution time which can surely affect the optimality of completion time. This paper provides a method to find the best combination of performers based on execution time of process instances by employing Bayesian Network (BN). We also determine the most critical variable that can impact on the overall system. We believe that this approach can assist business entities that want to improve their BP performance.