Polygonization of Point Clusters through Cluster Boundary Extraction for Geographical Data Mining
Interpretability and usability of clustering results are of fundamental importance. A linear time method for transforming point clusters into polygons is explored. This method automatically translates a point data layer into a space filling layer where clusters are identified as some of the resulting regions. The method is based on robustly identifying cluster boundaries in point data. The cluster polygonization process analyses the distribution of intra-cluster edges and the distribution of inter-cluster edges in Delaunay Triangulations. It approximates shapes of clusters and suggests polygons of clusters. The method can then be applied to display choropleth maps of point data without a reference map or to identify associations in the spatial dimension for geographical data mining.
Advances in Spatial Data Handling: 10th International Symposium on Spatial Data Handling