计算机集成制造系统 ›› 2016, Vol. 22 ›› Issue (第2期): 381-394.DOI: 10.13196/j.cims.2016.02.011

• 产品创新开发技术 • 上一篇    下一篇

用于面向过程的基于实例推理的语义工作流相似性算法

孙晋永1,2,古天龙2,闻立杰3,钱俊彦2   

  1. 1.西安电子科技大学计算机学院
    2.桂林电子科技大学广西可信软件重点实验室
    3.清华大学软件学院
  • 出版日期:2016-02-29 发布日期:2016-02-29
  • 基金资助:
    国家自然科学基金资助项目(61262030,61363030,61562015);桂林电子科技大学创新团队资助项目;广西高等学校高水平创新团队及卓越学者计划资助项目;广西可信软件重点实验室研究课题资助项目(KX201412)。

Similarity algorithm for semantic workflows used in process-oriented case-based reasoning

  • Online:2016-02-29 Published:2016-02-29
  • Supported by:
    Project supported by the National Natural Science Foundation,China(No.61262030,61363030,61562015),the Program for Innovative Research Team of Guilin University of Electronic Technology,China,the High Level of Innovation Team of Colleges and Universities in Guangxi and Outstanding Scholars Program Funding,China,and the Guangxi Key Laboratory of Trusted Software,China(No.KX201412).

摘要: 为了对同时包含控制流和数据流元素的语义工作流进行相似性度量,定义有代表性的迹集合表达语义工作流行为,提出获取有代表性的迹集合的方法。基于迹比对方法计算有代表性的迹的相似性,使用A*算法结合剪枝策略确定有代表性的迹集合的最优匹配,进而完成语义工作流的相似性计算。实验评估表明,该方法在语义工作流检索性能评价上具有良好的效果。

关键词: 面向过程CBR, 语义工作流, 有代表性的迹, 迹比对, 相似性

Abstract: To measure similarities between semantic workflows with control-flow and data-flow elements,the representative traces set was defined to express semantic workflow's behavior,and the method to gain the representative traces set was also introduced.Meanwhile,the representative traces similarity metric was put forward with trace alignment.The optimal matching between representative traces sets was determined by A* algorithm with pruning strategy,based on which semantic workflows similarity metric was proposed.The experiment showed the better effect of proposed method on retrieval performance evaluation of semantic workflows.

Key words: process-oriented case-based reasoning, semantic workflows, representative traces, trace alignment, similarity

中图分类号: