Semantic selection of georeferencing services for urban management

Abstract

Geocoding has become one of the most popular on-line services. Nowadays, there exist many Web Services providing geocoding functionality which differ not only in technological aspects (interface, invocation style, etc.) or terms of use, but also in type of geographic information provided and spatial data quality. Currently, there is no problem to find geocoding providers but to choose the proper one, which is determined by user’s requirements. The public administration is responsible for providing the official and appropriate information for citizens or deal with risk and health management issues; therefore, urban management systems require geocoding services of high quality in terms of quality of service (QoS) and spatial data. However, the present Web geocoding public market is dominated by geocoding services for average users, i.e. users that can accept low QoS or are not interested in the lineage of data. In this situation, a compound geocoder can join data from several services of geographic information and provide services oriented to the users that demand high quality, for example urban management systems. This paper presents an architectural approach for adaptive compound geocoding Web services built on diverse Web services of geographic information, such as gazetteers, cadastral services and address geocoding services. The proposed architecture is characterized by extensibility and adaptivity thanks to application of the ontologies (Administrative Unit Applied Ontology of Spain and Service Characteristic Ontology) and advances in Semantic Web related to strategies for source selection and result data evaluation, which are introduced in some details. As an example of real use case, it is presented the implementation of address geocoding compound Web service whose instances are applied in diverse tasks in urban management systems of Zaragoza City council in Spain.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 4 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.