On Measuring Process Model Similarity Based on High-Level Change Operations

For various applications there is the need to compare the similarity between two process models. For example, given the as-is and to-be models of a particular business process, we would like to know how much they differ from each other and how we can efficiently transform the as-is to the to-be model; or given a running process instance and its original process schema, we might be interested in the deviations between them (e.g. due to ad-hoc changes at instance level). Respective considerations can be useful, for example, to minimize the efforts for propagating the schema changes to other process instances as well. All these scenarios require a method to measure the similarity or distance between two process models based on the efforts for transforming the one into the other. In this paper, we provide an approach using digital logic to evaluate the distance and similarity between two process models based on high-level change operations (e.g. to add, delete or move activities). In this way, we can not only guarantee that model transformation results in a sound process model, but also ensure that related efforts are minimized.