Using Memetic Algorithm For Matching Process Models

To support the analysis, redesign, and implementation projects in enterprises, it's necessary to identify the correspondences between different business process models. However, according to the process model matching contests, the effectiveness of state-of-the-art process model matchers is low, i.e., their alignment only contain a few existing and many irrelevant correspondences. To improve the quality of process model alignment, in this work, we propose a novel activity similarity measure to calculate the similarity between different activities, construct an optimal model for process model matching problem and design a Memetic Algorithm (MA) based process model matcher to determine the high quality process model alignments. The experimental result shows that our approach's performance significantly outperforms other EA based matchers and state-of-the-art process model matchers.