[ 1 ] [ 2 ] [ 3 ] [ 4 ] [ 5 ] [ 6 ] [ 7 ] [ 8 ] [ 9 ] [ 10 ] [ 11 ] [ 12 ] [ 13 ] [ 14 ]

Journal of Information Science and Engineering, Vol. 35 No. 6, pp. 1343-1363

Spatial Co-location Pattern Mining Based on Fuzzy Neighbor Relationship

1Department of Computer Science and Engineering
Yunnan University
Kunming, 650504 P.R. China

2School of Information Science and Technology
Chuxiong Normal University
Chuxiong, 675000 P.R. China
E-mail: meijiaow@cxtc.edu.cn; lz hwang@ynu.edu.cn; 1657611732@ynu.edu.cn

A co-location pattern is a subset of spatial objects whose instances are frequently located together in geography space. The traditional co-location mining algorithms treated the spatial proximity relationship between the instances as unanimous by binary logic, which weakened the accuracy and effectiveness of the results. In this paper, the co-location pattern mining based on fuzzy neighbor relationship is studied. Firstly, fuzzy neighbor relationship (FNR) is defined to measure the proximity level between instances, and then the fuzzy participation ratio and the fuzzy participation index are defined. Secondly, the algorithm for spatial co-location pattern mining based on FNR (CPFNR) is proposed by the basic idea of the Join-less algorithm. Moreover, optimizing strategy is adopted for the CPFNR algorithm. Finally, the effectiveness of the CPFNR algorithm is verified by experiments on the real datasets, and the performance of our algorithm is evaluated on the synthetic datasets.

Keywords: data mining, fuzzy set, spatial co-location pattern, fuzzy neighbor relationship, fuzzy participation index

  Retrieve PDF document (JISE_201906_11.pdf)