From Business Process Model and Notation to Stochastic Automata Network.
Résumé
The qualitative and quantitative analysis of operational processes recently started to receive special attention with the business process management systems. But the Business Process Model and Notation (BPMN), the standard representation of business processes, is not appropriate to support the analysis phase. Most of the works proposing mappings from BPMN to formal languages aim model verification, but few are directed to quantitative analysis. In this work, we state that a well-defined BPMN Process diagram can originate a Stochastic Automata Network (SAN) - a compositionally built stochastic model. More than support verification, SAN provides a numerical evaluation of processes' performance. SAN attenuates the state-space explosion problem associated with other Markovian formalisms and is used to model large/complex systems. The main contribution of this work is an algorithm that converts BPMN diagrams to SAN. This conversion is the first step to build complete performance evaluation models of business processes.