›› 2017, Vol. 2 ›› Issue (12): 27-31.

• 数学与计算机科学 • 上一篇    下一篇

空间co-location模式挖掘表实例的计算方法

  

  1. (大理大学数学与计算机学院,云南大理671003)
  • 收稿日期:2017-08-13 出版日期:2017-12-15 发布日期:2017-12-15
  • 作者简介:曾新,讲师,主要从事数据挖掘、计算机应用技术研究.
  • 基金资助:
    国家自然科学基金资助项目(71462001);云南省科技厅应用基础青年基金资助项目(2016FD071);云南省教育厅应用基础研究基金资助项目(2016ZZX192)

A Calculation Method of Table Instances in Spatial Co-location Pattern Mining

  1. (College of Mathematics and Computer, Dali University, Dali, Yunnan 671003, China)
  • Received:2017-08-13 Online:2017-12-15 Published:2017-12-15

摘要: 为了优化候选模式表实例的计算开销,基于前缀模式的候选模式表实例的计算方法被提出,并分别将候选模式的前缀模式表实例存储于数据库(join-based-D)和字典(join-based-Z)中,进行候选模式表实例的计算。实验证明:基于前缀模式的候选模式表实例的计算方法join-based-Z的运行效率要优于join-based-D的运行效率。

关键词: co-location模式, 前缀模式, 表实例, 候选模式

Abstract: To optimize the computational cost of the table instances of candidate patterns, a computational method based on prefix patterns is proposed, and the table instances of candidate patterns is computed when the prefix patterns are stored in database(joinbased-D)and in dictionary(join-based-Z). The experiment shows that the computational method of join-based Z is more efficient than join-based-D.

Key words: co-location pattern, prefix pattern, table instance, candidate pattern

中图分类号: