Abstract:
The open data movement has resulted in many datasets on the web to be freely available for anyone to freely access, use, modify and share for any purpose (subject, at most, to requirements that preserve provenance and openness). The Linked Open Data (LOD) cloud is an example of such an open data repository of attribute data in the form of billions of Resource Description Framework triples on the Web. Visualising such open data in thematic maps provides a powerful spatial analysis tool for planning and decision-making. In this research, several styles of creating web thematic maps by integrating the attributes from the LOD cloud with geometry in a spatial database server were investigated and evaluated. Requirements for a specialised geospatial web service that combines linked data with geospatial data to create thematic maps were specified. Standard technologies were used, motivated by the widespread deployment of standardised web map services in the geospatial community and the widespread publication of alphanumeric data (by statistical agencies) in the LOD cloud. A specialisation of an Open Geospatial Consortium Web Map Service (WMS) that creates web thematic maps by integrating alphanumeric linked data from the LOD cloud with geometry in a spatial database server was conceptualised. Three integration styles (referred to as design options) for this specialised web service were designed and implemented. The first style integrates linked data with spatial data by an importer. The second and third styles use a middleware and extension of a spatial database server respectively to integrate linked data with spatial data. In each of the three styles, attributes are retrieved from the LOD cloud through semantic queries and only the results of the semantic query are visualised on the thematic map. In this way, the benefits of semantic queries are exploited in the Semantic Web itself and the WMS mapping capabilities are used to visualise the semantic query results on a thematic map by integrating these with geospatial data. The three integration styles are critically evaluated against the specified requirements. This research contributes to understanding the pros and cons of incorporating semantic (linked) data models into standard geospatial web service models to create cartographic products (web thematic maps). This research contributes to bridging the gap between linked data and web thematic mapping.
