Load triples into the following graph:
<http://lib.uchicago.edu/uchicagonode>
Then you can write SPARQL queries like this and confine the results to UChicagoNode triples regardless of whatever else might be in there:
prefix edm: <http://www.europeana.eu/schemas/edm/> prefix ore: <http://www.openarchives.org/ore/terms/> prefix uchicago: <http://lib.uchicago.edu/> select distinct ?collection from <http://lib.uchicago.edu/uchicagonode> where { { ?a a ore:Aggregation } . { ?a uchicago:inCollection ?collection } . { ?a edm:aggregatedCHO ?s} . }
There was an error while loading. Please reload this page.