| Resumo: | Business Process Modeling (BPM) is a systems engineering activity where we rep- resent the processes of an enterprise, so they can be shared, understood and improved. Despite the set of innovative tools for BPM modelling that exist in the market, they allow modelers to introduce errors during the modelling process. As there is no idea which errors the tools do not detect, what are the most recurrent errors and how could this prob- lem be mitigated, this dissertation presents a study and a proposal to help solving this problem. Firstly, a tool survey was developed to describe the state of the practice on the ability of Modelling Tools to validate BPMN2 models and determine the most recurrent defects introduced by BPMN modellers. Secondly, based on an empirical study using the QUASAR validator we provide evidence on its ability to validate a set of well-formedness rules and best practices and therefore detect errors in BPMN2 Models. Finally, we want to understand if this metamodelling-based validation facility can be used to prevent intro- ducing modelling errors, while speeding up the learning curve.
|
|---|