Method of Knowledge Representation on Spatial Classification

Spatial data mining is a highly demanding field because very large amounts of spatial data have been collected in various applications, ranging from remote sensing (RS), to geographical information system (GIS), computer cartography, environmental assessment and planning, etc. Classification is a data mining technique where the data stored in a database is analyzed in order to find rules that describe the partition of the database into a given set of classer. Knowledge Representation developed as a branch of artificial intelligence. As a result, the AI design techniques have converged with techniques from other fields, especially database and object-oriented system. In this paper, an efficient knowledge representation for classification of spatial data is proposed and studied. Our approach to spatial classification is based on both non-spatial properties of the classified objects and attributes, predicates and functions describing spatial relations between classified objects and other features located in the spatial proximity of the classified objects. We address issues regarding classification of spatial data and concentrate on building decision trees for the classification of such data. Furthermore, we produce rules that divide set of classified objects into a number of groups, where objects in each group belong mostly to a single class, and design a simple and convenient structure of knowledge base, which is based on relational data base. Finally, we visually represent the spatial classification result by using thematic map which showed the effectiveness of the proposed method.