Ontological modelling of geographical relationships for map generalization

Abstract

Understanding and encoding the roles that map features play in landscapes can assist making maps for particular purposes and for generalizing those maps appropriately. This paper explores how enumerating and classifying toponyms (place names) can enhance cartographic databases to allow applications to consider contextual relationships when choosing what to render or not and what symbolization rules to apply. A case study is presented of a coastal region in the U.S. to illustrate how ontologies can be formalized to support map display and generalization at a conceptual level that goes beyond traditional feature coding standards. While this might seem to some a return to a “capes and bays” geography (rote itemization of isolated geographic facts), it is in fact an effort to integrate toponyms into formalized knowledge framework expressed in a standardized, content- neutral language that can be shared by software anywhere on the World Wide Web.