Abstract: For large-scale complex organizations,it is a key problem in business process management that how to quickly find the appropriate process model what the user needs in the model repositories.Aiming at this problem,a process retrieval method based on performance constraints was proposed.In this method,the business process was transformed into the Breadth First Topology Sequence (BFTS) and the similarity values of the process modeles was calculated,thus s the candidate set was determined to reduce the retrieval space.The three performances of running time,cost and human resources of the candidate business process were calculated,and the optimal process model was selected.Comparative experiments verified that BFTS's operation efficiency and retrieval accuracy were higher than traditional greedy algorithm.

Key words: business process, retrieval, performance constraint, breadth first search sequence, graph edit distance

摘要： 大规模复杂组织如何快速地在模型库中找到用户所需的合适的过程模型是业务过程管理的一个关键性问题。针对该问题,提出了基于性能约束的过程模型检索方法。该方法首先将业务过程转换成广度优先拓扑序列,计算过程模型的相似度值,确定候选集,缩小了检索空间;然后对候选业务过程的运行时间、花费和人力资源三个性能分别进行计算,选出最优过程模型。最后,通过对比实验验证了所提方法的有效性。

关键词: 业务过程, 检索, 性能约束, 广度优先搜索序列, 图编辑距离