Object-oriented spatial analysis: Set-based exploratory analysis of the relations among the distributions of spatial objects

This paper develops a new exploratory method for analyzing the relations among distributions of spatial objects. Existing methods usually focus on the distribution of one or two types of spatial objects. This paper, on the other hand, deals with the relations among more than two distributions of spatial objects. In addition, the method is applicable independent of the type of spatial objects. These advantages assure a wide applicability and flexibility of the method. Analysis starts with the evaluation of local spatial relations among objects. This paper then develops a computational algorithm for clustering distributions based on a specific spatial relationship such as spatial proximity. The result is visualized by a graph-based representation and evaluated by numerical measures. The method is applied to the analysis of the distributions of commercial facilities in Chiba City, Japan. Technical soundness of the method is discussed as well as empirical findings.