Development of Metrics for Evaluating Decomposed Process Models based on Wand and Weber's Good Decomposition Model

Business process modelling is an important task in business transformation initiatives. Process models visualize the working procedures of a company and pinpoint the way in which business value is created. Based on process models, functional requirements on IS are derived and decisions on IS investments are made, for instance. However, in case process models become too large, employees will hardly understand them, which restricts the potential benefits associated with business process modelling. Therefore, the decomposition of process models is a means of reducing their complexity by delineating corresponding subprocess models. However, there are few commonly accepted approaches for decomposing process models only and the properties that characterize a well-done decomposition are rather unclear. We thus revert to the good decomposition model of Wand and Weber, which was established for decomposing IS, as a means to judge the quality of decomposed process models. The present study develops metrics for evaluating decomposed process models in the eEPC notation against the good decomposition model of Wand and Weber. An application of the metrics to a process model from a cooperation project shows that the metrics provide a helpful way of objectively assessing the quality of decompositions in process modelling by using the good decomposition model.