Research And Application About Grid Density-based Clustering Methods In GIS

Spatial Data Mining or Spatial Knowledge Discovering refers to picking up connotative knowledge, rules or significative modes from spatial database. This technology has great significance in understanding spatial data and picking up relations between spatial data and non-spatial data. Along with the development of Spatial Information System in which GIS ( Geography Information System ) is an important example a mass of spatial data is accumulated, so spatial data mining has become a crucial subject.Clustering is an important problem in the field of data mining. Clustering is a task to divide samples into some groups so samples of same group are more similar and those of different groups are less similar. Spatial clustering is to find clusters in spatial data, it inherits the characters of general clustering which focus on the process to find groups of space distributed samples. This article firstly research on the characters of distribution of spatial samples, introduces the dualism of distribution in geography space and the characters of coordinate-based spatial clustering in 2D space. A summary of spatial clustering in GIS is give out in the end.Secondly, this article researches on the existing clustering methods, especially on the Grid Density-based methods, both advantages and disadvantages are introduced. Then a new approach is presented, which is developed from traditional grid density-based methods. New methods can exactly and efficiently divide the objects in geography space into some density-based regions and divide each of these regions into some sub-regions at the same time which are approximate to circle and have a center. Relation between two kinds of regions are presented as a two-level frame. A method of parameters commending is also introduced for the conditions when there's no user's parameters.To deal with the spatial obstacles, this article introduces some existing means and gives out a means for Grid Density-based methods by a special way to turn the obstacles into abstract forms.Finally, the new methods and two other methods were integrated into the GIS and the whole configuration of Intelligent GIS system is introduced.