A folksonomy-ontology-based digital gazetteer service

At present, many digital gazetteer systems manage names and types of places. However, information of names and types is just a part of geographical knowledge, and types are relative nonflexible. Meanwhile, as free-tagging classification is widely used in web 2.0 implementations, folksonomy has raised various issues regarding information retrieval, though some management problems emerge when the quantity of inorganized tags raise to be large. In this paper, we describe an approach that embeds folksonomy and ontology into Digital Gazetteer Service in order to give a smarter Digital Gazetteer Service (FODGS), and to enrich geographical information retrieval capabilities. To present relations between tags and avoid problems raised by inorganized free-tagging, we develop a new method to manager those tags - we write a set of modular ontologies. For the purpose of make queries over all the data above, we store the data in a triple-store, which offers SPARQL support. Finally, this paper also introduces an application implementation - a prototype system of FODGS was built with a XML-based interface.