基于任务最短跟随距离矩阵的流程模型行为相似性算法  

SSDT matrix-based behavioral similarity algorithm for process models

作　　者：汪抒浩[1] 闻立杰[1] 魏代森[2] 王建民[1] 闫志强[1,3] 

机构地区：[1]清华大学软件学院,北京100084 [2]山东大学计算机科学与技术学院,山东济南250101 [3]首都经济贸易大学信息管理与信息系统系,北京100070

出　　处：《计算机集成制造系统》2013年第8期1822-1831,共10页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(61003099);国家863计划资助项目(2012AA040904);教育部-中国移动科研基金资助项目(MCM20123011)~~

摘　　要：目前的流程模型行为相似性算法有着不同程度的缺陷,为此首次提出任务最短跟随距离矩阵的概念,以很好地体现流程模型的行为特征,并给出了基于Petri网完全有限前缀的高效计算方法。基于该矩阵提出的行为相似性算法能完整支持流程模型各类结构。与现有算法的对比表明,该算法更符合流程模型行为相似性算法应该具有的各项性质。Aiming at the various degree of defects existed in behavioral similarity algorithm between process models, the concept of Shortest Succession Distance between Tasks （SSDT） matrix was introduced, which could represent the behavioral characteristics of process models well. Meanwhile, an efficient computing method based on complete finite prefix of Petri nets was proposed. The behavioral similarity algorithm based on SSDT matrix was put forward to sup- port all possible complex structures of process models. Compared with current approaches, the proposed algorithm could meet all the prospective properties better.

关 键 词：流程模型 行为相似性 任务最短跟随距离矩阵 算法 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]