• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2014, Vol. 36 ›› Issue (03): 536-540.

• 论文 • 上一篇    下一篇

寻找地震相关区域的词频共现分析方法

胡明生1,2,贾志娟1,吉晓宇1,洪流2   

  1. (1.郑州师范学院软件研究所,河南 郑州 450044;2.华中科技大学系统工程研究所,湖北 武汉 430074)
  • 收稿日期:2012-08-30 修回日期:2013-01-14 出版日期:2014-03-25 发布日期:2014-03-25
  • 基金资助:

    国家自然科学基金资助项目(U1204703/G011202);河南省重点科技攻关项目(122102310004);郑州市创新型科技人才队伍建设工程(10LJRC190)

A method of co-occurrence frequency analysis to
find out correlations among earthquake areas               

HU Mingsheng1,2,JIA Zhijuan1,JI Xiaoyu1,HONG Liu2   

  1. (1.Institute of Software,Zhengzhou Normal University,Zhengzhou 450044;
    2.Institute of Systems Engineering,Huazhong University of Science and  Technology,Wuhan 430074,China)
  • Received:2012-08-30 Revised:2013-01-14 Online:2014-03-25 Published:2014-03-25

摘要:

地震的地区相关性反映出一定地理位置上地震发生的规律性,寻找中国震区之间的相关性规律可以提前预测地震发生的区域。以中国历史地震灾害数据库为平台搭建共现分析模型,首先将地震区域以熟知地名进行划分,在此基础上改进原有Salton模型,提出一种基于修正Salton指数的共现分析方法,使用修正后的共现分析方法求解地震区域间的关联度。实验表明,该修正Salton模型能更准确地反映关联度。

关键词: 地震区域, 词频共现, Salton指数, 相关性

Abstract:

Correlations among earthquake areas reflect the regularity of earthquakes happening in certain areas. Looking for correlations among earthquakes can forecast the earthquake areas before the disaster comes. Based on DCHED (Database of Chinese Historical Earthquake Disasters), cooccurrence analysis model is established. Firstly, earthquake areas are divided according to the wellknown names. Secondly, the traditional Salton model is improved. And, based on the improved Salton index, a method of cooccurrence analysis is proposed in order to solve problems of correlations among earthquake areas. Finally, experiments are performed to prove that the improved Salton model can reflect correlations among earthquake areas more accurately.

Key words: earthquake areas;co-occurrence;Salton index;correlation