Abstract
The business performance of an enterprise tightly depends on the quality of its business process model (BPM). This dependence prompted several propositions to improve quality sub-characteristics (e.g. modifiability and reusability) of a BPM through transformation operations to change the internal structure of the model while preserving its external behaviour. Each transformation may improve certain metrics related to one quality sub characteristic while degrading others. Consequently, one challenge of this model transformation-based quality improvement approach is how to identify the application order of the transformations to derive the "best" quality model. This paper proposes a local optimization-based, heuristic method to decide on the application order of the transformations to produce the best quality BPM. The method is guided by both the perspectives, and the impact of each transformation on the quality metrics pertinent to the perspectives as well as the quality sub characteristics of interest to the designer. The method's and an experimental evaluation are presented.