http://www-sop.inria.fr/edelweiss/fabien/docs/w3c/rdfsource/rdfsource.html<br />http://purl.org/dc/elements/1.1/creator<br...
preaching to the choir ?!?<br />
RDF needs necktiesseems more controversial<br />
provenance, source, origin, author,…<br />creation date, age, use-by date,…<br />accuracy, authentication, certification, ...
wecanquery…<br />PREFIX dc: <http://purl.org/dc/elements/1.1/><br />PREFIX foaf: <http://xmlns.com/foaf/0.1/><br />SELECT ...
wecannotname…<br /><ul><li>reification quads do not imply the triples, no link
rdf:ID still at triple level and still no link
document URI, ad-hoc, bad practice</li></ul>… naming is currently done outside RDF<br />#me<br /><br />subject<br />#name...
named graphs, provenance and trust<br />[Carroll, Bizer, Hayes, Stickler, 2005]<br />communicate assertional intent and si...
nested typed graphs<br />[Sowa, 84] [Mugnier & Chein, 96] [Chein & Mugnier, 08]<br />“A typed nested conceptual graph and ...
RDF/XML syntax extension<br />[Gandon, Bottolier, Corby, Durville , 2007]<br /><rdf:Descriptionrdf:about="http://www- sop....
before<br />http://.../rdfsource.html<br />dc:title<br />dc:creator<br />http://...#me<br />RDF Source<br />foaf:mbox<br /...
after<br />http://.../rdfsource.html<br />dc:title<br />http://www.w3.org/biblio<br />dc:creator<br />http://www.w3.org/bi...
e.g. social tagging<br />sioc:has_creator<br />http://.../tag23<br />#buttersg88<br />dc:date<br />11/03/2008<br />“nature...
e.g. concurrent hypotheses <br />context 2<br />context 1<br />context 3<br />context 6<br />context 4<br />context 5<br /...
e.g. tracing and backtracking<br />source 2<br />source 1<br />source 1,2,3,4<br />source1,2,3<br />service 3<br />service...
of 16

Name That Graph !

presentation on named graphs at the RDF next step workshop
Published on: Mar 3, 2016
Published in: Technology      Education      
Source: www.slideshare.net


Transcripts - Name That Graph !

  • 1. http://www-sop.inria.fr/edelweiss/fabien/docs/w3c/rdfsource/rdfsource.html<br />http://purl.org/dc/elements/1.1/creator<br />http://purl.org/dc/elements/1.1/title<br />http://ns.inria.fr/fabien.gandon/foaf#me<br />RDF Source<br />http://xmlns.com/foaf/0.1/name<br />http://www.w3.org/1999/02/22-rdf-syntax-ns#type<br />http://xmlns.com/foaf/0.1/mbox<br />mailto:fgandon@inria.fr<br />Fabien Gandon<br />http://xmlns.com/foaf/0.1/Person<br />http://<br />namethat graph!<br />Fabien Gandon, Olivier Corby, ,<br />
  • 2. preaching to the choir ?!?<br />
  • 3. RDF needs necktiesseems more controversial<br />
  • 4. provenance, source, origin, author,…<br />creation date, age, use-by date,…<br />accuracy, authentication, certification, validity, …<br />context, assumption, modality, …<br />dependencies, disjunction, negation, …<br />access rights, copyrights, …<br />location, society, culture, …<br />...informational context<br />
  • 5. wecanquery…<br />PREFIX dc: <http://purl.org/dc/elements/1.1/><br />PREFIX foaf: <http://xmlns.com/foaf/0.1/><br />SELECT *<br />WHERE<br />{<br />GRAPH ?gphdoc { ?docdc:title ?title .<br />?docdc:creator ?author }<br />GRAPH ?gphauthor { ?authorfoaf:name ?name }<br />}<br />
  • 6. wecannotname…<br /><ul><li>reification quads do not imply the triples, no link
  • 7. rdf:ID still at triple level and still no link
  • 8. document URI, ad-hoc, bad practice</li></ul>… naming is currently done outside RDF<br />#me<br /><br />subject<br />#name<br />predicate<br />#S<br />#me<br />#name<br />"fabien"<br />object<br />"fabien"<br />#me<br /><br />subject<br />#name<br />predicate<br />#me<br />#S<br />"fabien"<br />#name<br />rdf:ID="S"<br />object<br />"fabien"<br />≠<br />
  • 9. named graphs, provenance and trust<br />[Carroll, Bizer, Hayes, Stickler, 2005]<br />communicate assertional intent and sign assertions<br />embody social acts with some record<br />name with URIs & graph RDFS class<br />:G1 {_:Monica ex:name "Monica Murphy" .<br /> _:Monica ex:email <mailto:monica@murp.hy> .<br /> :G1 pr:disallowedUsagepr:Marketing }<br />:G2 { :G1 ex:author :Chris .<br /> :G1 ex:date "2003-09-03"ˆˆxsd:date }<br />
  • 10. nested typed graphs<br />[Sowa, 84] [Mugnier & Chein, 96] [Chein & Mugnier, 08]<br />“A typed nested conceptual graph and a coreference link (dashed line)” [Genest & Salvat, 98]<br />
  • 11. RDF/XML syntax extension<br />[Gandon, Bottolier, Corby, Durville , 2007]<br /><rdf:Descriptionrdf:about="http://www- sop.inria.fr/edelweiss/fabien/docs/w3c/rdfsource/rdfsource.html"cos:graph="http://www.w3.org/biblio" ><br /> <dc:title>RDF Source</dc:title><br /> <dc:creator><br /> <foaf:Personrdf:about="http://ns.inria.fr/fabien.gandon/foaf#me" <br />cos:graph="http://www.inria.fr/hr" ><br /> <foaf:name>FabienGandon</foaf:name><br /> <foaf:mboxrdf:resource="mailto:fgandon@inria.fr"/><br /> </foaf:Person><br /></dc:creator><br /></rdf:Description><br />
  • 12. before<br />http://.../rdfsource.html<br />dc:title<br />dc:creator<br />http://...#me<br />RDF Source<br />foaf:mbox<br />rdf:type<br />foaf:name<br />mailto:fgandon@inria.fr<br />Fabien Gandon<br />foaf:Person<br />
  • 13. after<br />http://.../rdfsource.html<br />dc:title<br />http://www.w3.org/biblio<br />dc:creator<br />http://www.w3.org/biblio<br />http://...#me<br />RDF Source<br />foaf:mbox<br />rdf:type<br />foaf:name<br />http://www.inria.fr/hr<br />http://www.inria.fr/hr<br />http://www.inria.fr/hr<br />mailto:fgandon@inria.fr<br />Fabien Gandon<br />foaf:Person<br />
  • 14. e.g. social tagging<br />sioc:has_creator<br />http://.../tag23<br />#buttersg88<br />dc:date<br />11/03/2008<br />“nature”<br />sioc:container<br />http://www.flickr.com<br />
  • 15. e.g. concurrent hypotheses <br />context 2<br />context 1<br />context 3<br />context 6<br />context 4<br />context 5<br /><br /><br />
  • 16. e.g. tracing and backtracking<br />source 2<br />source 1<br />source 1,2,3,4<br />source1,2,3<br />service 3<br />service 4<br />
  • 17. e.g.…<br />temporal aspects in querying and reasoning.<br />context of knowledge extraction (NLP).<br />capture of points of view.<br />distributed storage and query routing.<br />security<br />etc.<br />
  • 18. names, plllleaaase,@TODO blank nodes, one triple in several graphs, backward compatibility, rdf:Graph & rdf:GraphStore// comments - RDF 1.1 not RDF 2.0 - Follow your nose for RDF and named graphs - RDF meta-model must remain domain independent to standardize a best practice there must be a practice in the first place<br />Fabien Gandon, Olivier Corby, ,<br />