计算机应用与软件Issue(12)：185-189,5.DOI:10.3969/j.issn.1000-386x.2013.12.049
集成同位模式的空间关联规则挖掘方法
MINING METHOD OF SPATIAL ASSOCIATION RULE INTEGRATING CO-LOCATION PATTERNS
摘要
Abstract
In spatial data mining , to use traditional frequent item-set mining method will lead to duplicated counting of spatial entities;moreover , it does not take into account that the relations between spatial entities and the correlation of spatial entities and their surrounding environments will produce a large amount of irrelevant spatial information .In order to overcome the deficiency , in the paper we propose the model and the algorithm of spatial association rule mining which integrates the spatial co -location patterns mining .The idea of "the first law of geography"and "the biocoenosis construction"are introduced .Firstly, it carries out the optimised partition on spatial entities features distri-bution and the discretised process on continuous space distribution .Secondly , it eliminates the redundant spatial information , and uses spatial co-location algorithm to mine the implicit relationships of different spatial entities features .The spatial attributes , no-spatial attributes and spatial relationships of the entities are employed to construct the spatial transaction database .Lastly, the mining of spatial association rules is conducted in spatial transaction database .Experimental results show that the model and the algorithm are effective , the correlation between the spatial entities features is considered .Many significant spatial association rules can be discovered from the constructed spatial transaction database after the elimination of the redundant information and the discretised processing on the continuous space distribution .关键词
空间关联规则/生物群落原理/数据挖掘/空间同位模式Key words
Spatial association rule/Biocoenosis principle/Data mining/Spatial co-location pattern分类
信息技术与安全科学引用本文复制引用
向俊,王静..集成同位模式的空间关联规则挖掘方法[J].计算机应用与软件,2013,(12):185-189,5.基金项目
广西教育厅科研项目（201106LX868，  （）
201106LX864）；广西广播电视大学中青年项目（2012ZQN06）。 （）
