Structural Similarity Assessment of Business Process Graph Using GED-Greedy

Process automation in business process redesign is a critical issue. One of the problems is related to retrieving similar business process models (BPMs) stored in a repository based on implicit business process model queries extracted from user feedback. Previous studies in graph similarity focus on one-to-one business process model (BPM) similarity; however, this approach is considered time-consuming. This study aims to identify the possibility of using the graph approach and Graph Edit Distance Greedy (GED-Greedy) as structural similarity methods for measuring similarity between BPM queries and BPMs in the repository. The structure of a BPM can be represented as the inter-structure and intra-structure of a graph. The structural similarity value of the two graphs’ BPM final result is 0.91. It means that the proposed methods from the case study could be used for measuring structural similarity between BPMs.