Spatial Data Mining on Hierarchical Semantic Relation among Multi-scale Geographical Representations

Abstract

Multi-scale representations for geographical entities is one of research focuses in the field of multi-scale spatial database, Compared with the traditional GIS model, the essential modeling mechanism and strategies for building multi-scale data model are explored. Taken the road feature as an example, a detailed description of multi-representation model elements in the multi-scale model are provided, which includes abstract data types and attribute-value domains for geographical features at different scales, as well as the scale, semantic, and mapping hierarchical relations among multi-representations. Based on which, an appropriate multi-scale conceptual model is given in detail with multi-scale extended E-R method. At last, a case in point is given to implement the entities and relations abstracted from the multi-scale modeling technique strategies, which is helpful to build a convenient and visual multi-sale spatial information services platform for route design, route guide and route display.