Breadth First Search Sequence Based Method for Efficient Process Retrieval

With increasing improvement of the business process management (BPM) technology, large-scale business process repositories have been adapted widely. However, due to the explosion of the number of business processes, a large part of enterprises are confronting with the challenge on effective management of those massive processes. Usually, each business process is modeled as a process graph, and therefore most existing approaches are based on graph mining algorithms. This paper puts forward a new method, which first utilizes the breadth first search (BFS) algorithm to label the process model, and then calculates the similarity based on the matching distance. The experimental results show that our method is efficient enough for practical use, especially suitable for fuzzy retrieval.