Research On GIS-based Spatial Data Mining Method

Spatial Data Mining(SDM), an important branch of Data Mining, refers to extraction of implicit knowledge, spatial relationships, and other patterns which are not explicitly stored in Spatial Database. Geographic Information System (GIS) is a system based on Geospatial Database for geographic decision support. Introducing the concept, pattern and method of DM and KDD into the field of GIS, we can use the storage management and analysis function of GIS to make the limited data in GIS become the limitless knowledge. For doing so, we can effectively handle the huge data, advance the autoimmunization and intelligence level of the geography data analysis and offer powerful analysis tools for the globe change and sustainable development.This paper researches the spatial data mining based on GIS, commits to seek an effective way to discover knowledge from GIS. Beginning with the introduction of the background of SDM and GIS, the paper introduces the concept, methods and applications of spatial data mining and the definition, function and development direction of geographic information systems. Combining with the characteristics of spatial data mining and the functions of geographic information systems, the paper presents a GIS-based spatial data mining model, which gives full play to the advantages of both them. And based on the model, the methods of discovering spatial association rules and clustering from GIS are presented.Based on the above-mentioned study, a specific instance of geographical data and a simulation test are provided to verify the effectiveness of these two methods.