An approach for Mining Complex Spatial Dataset

摘要

Spatial data mining organizes by location what is interesting as such, specific features of spatial data mining (including observations that are not independent and spatial autocorrelation among the features) that preclude the use of general purpose data mining algorithms poses a serious challenge in the task of mining meaningful patterns from spatial systems. This creates the complexity that characterises complex spatial systems. Thus, the major challenge for a spatial data miner in trying to build a general complex spatial model would be; to be able to integrate the elements of these complex systems in a way that is optimally effective in any particular case. We have examined ways of creating explicit spatial model that represents an application of mining techniques capable of analysing data from a complex spatial system and then producing information that would be useful in various disciplines where spatial data form the basis of general interest.