特定空间对象同位模式挖掘算法研究

Algorithm of Mining the Specific Object Spatial Co-Location Pattern

空间同位模式挖掘研究主要以区域划分为基础,考虑对象实例两两之间的距离关系,这样挖掘出的同位模式是双向对称的。但区域的划分起止位置不确定,可能出现由于区域划分的不一致而得到不一样的空间同位模式结果。该文提出以指定对象为核心的空间同位模式挖掘,这样不必担心区域划分的起止位置对挖掘结果的影响,而且更能有针对性地发现特定空间对象与其它哪些对象具有空间同位关系。

The research concerning over Spatial data mining based mainly on partitioning of areas, in regard to the spatial distance between each object instance, Co-location pattern obtained will be symmetric. Yet the position where the partition begins and ends is indeterminate, therefore the Spatial Co-location pattern obtained could vary due to different partition. This paper proposed a specified object centered Spatial data mining method in case of inaccuracy caused by partitioning position, moreover, the finding of Spatial Co-location pattern between specific spatial object and the others can get more well-focused.