Trentino Government Linked Open Geodata: First Results

In this paper we report our work on publishing linked data of the Trentino Region in Italy following the Open Government Data (OGD) Initiative and in compliance with INSPIRE directive. Introduction: As part of the Semantic Geo-Catalogue (SGC) project [15, 16], the Autonomous Province of Trento (PAT) published its geographical data as linked data in RDF. Within the INSPIRE Directive1, it is required to have, among others, a resolvable unique identifier for each spatial object [3]. In the Linked Data [4], objects are named using resolvable HTTP URIs [5]. Therefore, the notion of Linked Data is in line with INSPIRE. In this paper we describe the methodology and design decisions we took to produce the linked data. Open Government Data: The OGD movement encourages governments to publish their data and content under a suitable license so that others can use, reuse and redistribute it through proper attribution. The Open Knowledge Foundation (OKF) community provides a list2 of such licenses. They are divided in two families: Open Data Commons3 (ODC) for data and Creative Commons4 (CC) for content. CC-Zero supports both data and content sharing and provides maximum flexibility in their use. Linked Data: In UK the INSPIRE Directive is followed. In particular, the UK government has decided to publish their data using open standards, e.g., RDF for representation [6], SPARQL Endpoint for exposing [4, 7], DCMI (Dublin Core Metadata Initiative) vocabulary for annotation [8] and GML (Geography Markup Language) for representing geographic features [9]. Essentially, the use of a SPARQL Endpoint for exposing data allows the Semantic Web search engines (for instance Sindice5, Swoogle6 and Watson7) to discover, crawl and index the RDF data which in turn helps increasing the visibility of the data itself. Ordnance Survey8, the national mapping 1 http://www.ec-gis.org/inspire/ 2 http://opendefinition.org/licenses 3 http://www.opendatacommons.org/ 4 http://creativecommons.org/ 5 http://www.sindice.com/ 6 http://swoogle.umbc.edu/ 7 http://watson.kmi.open.ac.uk/ 8 http://www.ordnancesurvey.co.uk agency in the UK, spearheaded the publishing of geospatial information as part of the Linked Data [10]. There are similar initiatives in EU. For example, in Spain the GeoLinked Data [12] initiative at the University Politecnica de Madrid has contributed bringing Spanish geographic and statistical information to the Linked Data. They have dealt with the data sources owned by the Spanish National Geographic Institute (IGN-E)9 and Spanish National Statistical Institute (INE)10. Their dataset is linked to GeoNames11 and DBPedia12. For the representation of the statistical (e.g., unemployment rate), geometrical (e.g., shape) and geo-positioning (e.g., geographical coordinates) information, Statistical Core Vocabulary (SCOVO)13, GML and WGS84 vocabularies were used, respectively. Trentino Linked Data: Under the regional deliberation n. 195/2012 the PAT formally decided to adopt CC-Zero to release 161 of its geographical datasets, such as streams and bicycle tracks. Data was available as shape files, while metadata (aligned with the ISO 19115 Core metadata elements) was available as XML files. To make them part of the Linked Data cloud [13], we converted both of them in RDF.