Linked GeoRef
1 like1,130 views
Linked GeoRef is an open dataset and API that aims to improve geographical annotation and finding of public sector information by integrating different geographical reference systems into a single linked data framework. It assigns unique identifiers (URIs) to codes from systems like AGS, NUTS, and Berlin's LOR, and links related codes to show hierarchical and equivalence relationships. This allows different systems to be used together and converted between each other. The dataset is hosted with SPARQL endpoint for lookup and conversion of codes between systems.
Linked GeoRef
- 1. Linked GeoRef An Open Dataset and API for Improving Geographical Annotation and Finding of Public Sector Information Knud M旦ller (Datalysator), Florian Marienfeld (Willh旦ft IT) 2nd Open Data Dialog, 19 November 2013, Berlin, Germany S trOR A LYns aAdTaining DAT olutio n web data s
- 2. The Challenge Most (Open (Government)) datasets refer to some location. Lets declare that in the metadata! http://www.opendata-hro.de D TA 2 LYSATOR web data solutions and training
- 3. Which Standard? There are plenty of options ISO authority: ISO 3166-1/2, Codes for the representation of names of countries and their subdivisions EU regulation: NUTS (Nomenclature of Units for Territorial Statistics) German elaboration: Amtlicher Gemeindeschl端ssel (AGS) / Amtlicher Regionalschl端ssel (ARS) Geometry: Geographical Bounding Boxes 3 D TA LYSATOR web data solutions and training
- 4. Best fit for Germany? AGS would work great From federal states down to boroughs: 05: Nordrhein-Westfalen 3: Regierungsbezirk K旦ln 82: Rhein-Sieg-Kreis 064: Gemeinde Swisttal 0538364 D TA 4 LYSATOR web data solutions and training
- 5. 053: Regierungsbezirk K旦ln 05382: Rhein-Sieg-Kreis 05: Nordrhein-Westfalen 5 Maps taken from Wikipedia 05382064: Gemeinde Swisttal D TA LYSATOR web data solutions and training
- 6. Problems with AGS only Germany not even all of it: no further detail for city-states like Berlin or Hamburg (no Neuk旦lln in AGS) city states have their own local systems, e.g. Lebensweltlich orientierte R辰ume (LOR) in Berlin D TA 6 LYSATOR web data solutions and training
- 7. Problems with others NUTS and ISO: not as detailed as AGS not at home in administration bounding boxes not precise enough coordinates not so easy to read as text Everybody likes their own approach best... D TA 7 LYSATOR web data solutions and training
- 8. Why Choose? Proposal: Linked Data solution combine all approaches in one dataset every code in every system gets unique identifier: a URI dereference it (surf to it) and find out what it means! codes are linked: what is the same? what is contained in what (hierarchy)? 8 D TA LYSATOR web data solutions and training
- 9. Linked GeoRef http://lgeoref.org/nuts/DE (Germany) http://lgeoref.org/nuts/DEA (North Rhine-Westphalia) http://lgeoref.org/ags/05 (North Rhine-Westphalia) http://lgeoref.org/ags/12 (Brandenburg) http://lgeoref.org/ags/05/3/82 (Rhein-Sieg-Kreis) http://lgeoref.org/berlin/lor/08 (Neuk旦lln) http://lgeoref.org/berlin/lor/02/02/02/06 (Graefekiez) D TA 9 LYSATOR web data solutions and training
- 10. Integrate Reference Systems Germany DE NRW Berlin DEA DE3 D TA 10 LYSATOR web data solutions and training
- 11. Integrate Reference Systems Germany DE Berlin NRW DEA 05 DE3 11 Reg.-Bez. K旦ln 05/3 Rhein-Sieg-Kreis 05/3/82 Swisttal D TA 05/3/82/064 11 LYSATOR web data solutions and training
- 12. Integrate Reference Systems Germany DE Berlin NRW DEA 05 FriedrichshainKreuzberg Reg.-Bez. K旦ln DE3 11 02 05/3 Rhein-Sieg-Kreis 05/3/82 02/02 02/02/02 08 Neuk旦lln Kreuzberg S端d Tempelhofer Vorstadt Swisttal 05/3/82/064 Graefekiez 12 02/02/02/08 D TA LYSATOR web data solutions and training
- 13. D TA 13 LYSATOR web data solutions and training
- 14. Usecase: govdata.de geo-referencing data map-based search and filtering hard to enforce one scheme in metadata Linked GeoRef could help: spatial-uri field in addition to spatial (coordinates) and spatial-text (human-readable): "spatial-uri": "http://lgeoref.org/berlin/lor/08" D TA 14 LYSATOR web data solutions and training
- 15. Usecase: Code Converter convert between different reference systems which AGS is this NUTS DEE? D TA 15 LYSATOR web data solutions and training
- 16. Collateral Usage central reference point for different code systems lookup for relations between codes (containment, identity, overlap, etc.) links to other widely-used reference datasets, e.g. DBpedia, GeoNames D TA 16 LYSATOR web data solutions and training
- 17. Internals data in various formats (mostly Excel) from various sources converted to RDF enriched and interlinked,Geonames) linked to external data (DBpedia, each reference system in its own namespace http://lgeoref.org/ags/ http://lgeoref.org/nuts/ http://lgeoref.org/berlin/lor/ etc. hosted in RDF store with SPARQL endpoint (dydra.com) simple Linked Data frontend with Pubby D TA 17 LYSATOR web data solutions and training
- 18. Take-home challenge: Open Data needs geographical context many useful geo reference systems (AGS, NUTS, ...) but: everybody likes a different one so: dont choose, use all of them together Linked GeoRef provides integrated, dereferenceable framework for this D TA 18 LYSATOR web data solutions and training
- 19. References NUTS: http://epp.eurostat.ec.europa.eu/portal/page/portal/ nuts_nomenclature/introduction AGS: http://de.wikipedia.org/wiki/Amtlicher_Gemeindeschl端ssel Dydra: http://dydra.com LOR: http://www.stadtentwicklung.berlin.de/planen/ basisdaten_stadtentwicklung/lor/ Pubby: https://github.com/cygri/pubby D TA 19 LYSATOR web data solutions and training