LinkedOpen Data forinspire: From3 to5 stargeospatialdata Francisco J. Lopez-Pellicer, AnetaJ. Florczyk, Javier Nogueras-Iso, Pedro R. Muro-Medrano and F. Javier Zarazaga-Soria INSPIRE Conference 2011, Edinburg, July 1, 2011
5 star Linked Data? Sir Tim Berners Lee (2010) This year, in order to encourage people- especially government data ownersalongtheroadtogoodlinkeddata, I havedeveloppedthisstarrating system Details@ http://www.w3.org/designissues/linkeddata.html Purpose of this talk Introduce this rating system in our context We vecome partway!!! Exemplified with datasets from Spain Presenttwo5 starlinkeddata sitesin SpainrelatedtoSDI nodes Presenta 5 starlinkeddata recipeforinspire data
1 Star MakeyourstuffavailableontheWeb (whateverformat) under an open license. MTN50 catographic grid Mapof MTN50 catographic grid Boundaries Gazetteer National Reference Geographic Equipment Public data, required attribution
2 Star Make it available as structured data (e.g., ESRI Shapefile instead of image portrayal of data). Mapof MTN50 catographic grid in png
3 Star Use non-proprietaryformats(e.g., WKT insteadof ESRI Shapefile, CSV instead of Access). MTN50 cartographic gridavailablein WKT and ESRI Shapefile Boundaries available only in ESRI Shapefile format Gazetteer available only in Access format
4 Star (no way?) Use URIstoidentifythings, so thatpeoplecan pointat your stuff (native use of RDF is not required) MTN50 grid not published as RDF GeoLinkedData.es initiative makes available this data as RDF
5 Star Link your data to other data to provide context.
5 Star+ Metadata For governement data, there should be metadata about the data itself(e.g. provenance, rights).
5 Star+ Metadata+ Data registry For governement data, their metadata should be available from an official registry(e.g. catalogue).
5 Star + Metadata + Data registry + Infrastructure Forgovernmentdata, thepreviousstepsrequirea coordinateseries of agreementsontechnologystandards, institutional arrangements, and policies.
Summary Level On the Web Metadata Data registry Infrastructure Datasets Open license(few or no limitations) Structured data (easy to use) Non-propietary format(no additional license restrains) URIsforthings(workwiththeWeb) RDF data model(open standardthatworkswiththeweb) Co-existence with other approaches(don t replace) Keep things simple (don t overengineering RDF) RDFlinks (workwiththeweb) Link to other approaches(don t replace) Simple vocs. such as dcat(keeping thing simple) Official registry(transparency) SPARQL (open standardthatworkswiththeweb) Agreements, Norms, Funds, Political support,
5 Star Linked Data examples from Spain - http://geo.linkeddata.es/ - http://datos.zaragoza.es/
Spanish NGI data @ http://geo.linkeddata.es/
Spanish NGI data @ http://geo.linkeddata.es/ Level On the Web Metadata Data registry Infrastructure Hydro topics of BCN200, BTN25, National Gazetteer Public data (Norm FOM/956/2008) Vector and alphanumeric data Vary, including gesopatial proprietary formats Developed and maintained by academia Simple vocabulary Developed by academia Do notlink tosdi resources Provenance vocabulary CKAN (no official) National Geographic Instute(IGN) support
Saragossa City Council data @ http://datos.zaragoza.es/
Saragossa City Council data @ http://datos.zaragoza.es/ Level On the Web Metadata Data registry Infrastructure Assorted content, including data from local SDI node Law: Local statute& Law37/2007 Web: ColorIURIS License Points of interest and events Vary, including geospatial proprietary formats Developed by industry, maintained by city council Encodedas GeoRSS/RSS 1.0 = RDF Clashes with existing Web admin practices(!) Developed by industry, maintained by city council Do notlink tosdi resources Dcat vocabulary Maintained by city council SPARQL City council, normative and funding support
Conclusions -Do sand don tsaboutlinkeddata -5 starrecipeforinspire data
Conclusions: do s and don ts about Linked Data Do Publish valuable data Pick persistent URIs for naming things Dereference URIs to representations URLs Put metadata giving license and provenance Use RDF formats for data transmission in addition Use SPARQL for data and metadata access Keep simple Integrate with existing systems Don t Publishallyourdata Publish outdated data Publish without explicit license Hidedata behindformsor applications Publish data only in propietary formats Waituntilyouhavea complete ontology Seek to replace existing systems
Conclusion: 5 star recipe for INSPIRE data Levels On the Web Metadata Topics INSPIRE datasets INSPIRE norms Transposed to Web licenses Vector data Representation in an open format [WE ARE ALREADY HERE!!!] A simplified representation in RDF Integrate with existing SDI geoportal/technologies Link to SDI resources Dcatvocabulary(crosswalkfromexistingmetadata) Data registry Enable the use of SPARQL to query existing SDI catalogue Infrastructure Linked Data as one of the agreements of the SDI [PREREQUISITE]
Francisco J. Lopez-Pellicer fjlopez@unizar.es http://iaaa.cps.unizar.es/ IAAA is currently a partner in the EuroGeoSource project http://www.eurogeosource.eu