A generic framework for flexible and data-aware business process engines

Abstract

Business Process Management can be applied to improve the efficiency, efficacy and transparency of its processes. This paper focusses on processes that can be characterized as loosely framed and knowledge-intensive (e.g. diagnosis and treatment of a patient in a hospital). Declarative process modeling languages that incorporate decision logic are well suited to model these processes and tools for the automatic discovery of such models from the data in the current IT-systems are starting to appear. However, the resulting models are often difficult for humans to understand, which makes it hard to implement them. Process engines can help to overcome this problem, as they can handle the complexity of the model and support the users to correctly execute it. Process engines are common ground for imperative models, but the offerings on the declarative side are more limited. This paper uses the design science methodology to develop a generic framework for flexible and data-aware process engines. The declarative model supports the direct flexibility needs of the process actor during execution, and it also enables a more general flexibility of the information system as changes to the model can be enacted immediately. An implementation of proposed framework for the DeciClare language was created, DeciClareEngine, and demonstrated with the realistic arm fracture case.