关键词: Petri网, 过程模型, 相似性度量, 模型存储

Abstract: With the improvement of business process and the change of market,many different versions of process models would appear which focus on the same business topic.Based on the structural similarity between those models,process model storage mechanisms were studied to improve model storage and to trace the change history easier for users.On the basis of using Petri net to represent the process model,greedy algorithm was applied to calculate the graph edit distance between different process models,and the basic edit operations and similarity formula were designed.Furthermore,five different process model storage solutions were proposed,and experiments were designed to make comparison.The results showed that the proposed storage could strike a balance between response time and storage space by combining with different demands of users.

Key words: Petri nets, process model, similarity measure, model storage