Connecting the SmithsonianAmerican Art Museum tothe Linked Data CloudPedro Szekely, Craig A. Knoblock, Fengyu Yang, Xuming...
The Smithsonian American ArtMuseum is a museum in Washington,D.C. which has one of the worldslargest and most inclusive co...
Big PicturePedro Szekely and Craig KnoblockUniversity of Southern California
ProblemSAAMDataWhat ontology to use?Structure mismatchesData consistency What to link to?100% precisionHow to enable museu...
Steps to Create Linked Data• Map data to RDF… select ontologies… define mappings• Link to external resources… identify the...
select ontologies
University of Southern California
ComplicatedMany irrelevant classesand propertiesIncompleteUniversity of Southern California
edm:ProvidedCHOaac:CulturalHeritageObjectdcterms:creatorore:Aggregationedm:EuropeanaAggregationcrm:E89_Propositional_Objec...
edm:ProvidedCHOaac:CulturalHeritageObjectskos:Conceptskos:Conceptedm:hasTypeskos:narrowerskos:prefLabelskos:prefLabelsaam:...
mapping the data tothe ontologieshow to enable museums to do this themselves?Pedro Szekely and Craig KnoblockUniversity of...
KarmaHierarchicalSourcesServicesModelKarmaTabularSourcesDatabase…Interactive tool for rapidly extracting, cleaning, transf...
specifying transformations andmapping to properties withKarmaPedro Szekely and Craig KnoblockUniversity of Southern Califo...
saam:person/2aac-ont:Person“George M. Aarons”aac-ont:variantNamerdf:typesaam:person/15“Alice Stanley Archeson”aac-ont:marr...
Pedro Szekely and Craig KnoblockUniversity of Southern Californiadownload the presentation to view the embedded video
mapping to objectproperties usingKarmaPedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern Californiadownload the presentation to view the embedded video
Evaluation of Data Mapping Using KarmaSAAM database8 tables29 columnsOntologies407 classes105 data properties229 object pr...
identifying andcurating linksPedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaMultiple “John Singer Sargent”ima:Person_John_Singer_Sarg...
Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaJohn Singer Sargentima:SaamPerson_John_Singer_Sargenta sa...
Linking “John Singer Sargent”saam:Person_4253owl:sameAs cb:Person_John_Singer_Sargent ;owl:sameAs dallas:Person_John_Singe...
IntuitionEstimate discrimination power of properties,e.g., of name, birth and death datesbirth date death date # of people...
Evaluation of Automatic LinkingPedro Szekely and Craig KnoblockUniversity of Southern CaliforniaSAAM names starting with “...
Results of Automatic LinkingGetty ULAN® 2,110Rijksmuseum 551Geonames 3,068DBPedia 2,194New York Times 70Pedro Szekely and ...
Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaCurating Links with Karma
Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaLinking with Karma
results of automated linking andinteractive curation recorded usingPROVPedro Szekely and Craig KnoblockUniversity of South...
deploymentPedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern California
Pedro Szekely and Craig KnoblockUniversity of Southern California
Related Work• Europeana• 17 million items, 1,500 institutions• Require exports in “Europeana” format• Amsterdam Museum, Mu...
Next Steps• Applications leveraging linked data• Virtual museum• Tools to create multimedia stories about art• Tools to fi...
Merci
Upcoming SlideShare
Loading in …5
×

Connecting the Smithsonian American Art Museum to the Linked Data Cloud

3,766 views

Published on

Slides for our "Connecting the Smithsonian American Art Museum to the Linked Data Cloud." paper presented at the 10th Extended Semantic Web Conference (ESWC), in Montpellier, May 2013. http://eswc-conferences.org/sites/default/files/papers2013/szekely.pdf

Published in: Technology, Education, Business
1 Comment
5 Likes
Statistics
Notes
No Downloads
Views
Total views
3,766
On SlideShare
0
From Embeds
0
Number of Embeds
521
Actions
Shares
0
Downloads
83
Comments
1
Likes
5
Embeds 0
No embeds

No notes for slide

Connecting the Smithsonian American Art Museum to the Linked Data Cloud

  1. 1. Connecting the SmithsonianAmerican Art Museum tothe Linked Data CloudPedro Szekely, Craig A. Knoblock, Fengyu Yang, Xuming Zhu,Eleanor E. Fink, Rachel Allen, and Georgina GoodlanderUniversity of Southern California, Los Angeles, California, USANanchang Hangkong University, Nanchang, ChinaSmithsonian American Art Museum, Washington, DC, USAhttp://www.isi.edu/integration/karma
  2. 2. The Smithsonian American ArtMuseum is a museum in Washington,D.C. which has one of the worldslargest and most inclusive collectionsof art, from the colonial period to thepresent, made in the United States.Wikipedia
  3. 3. Big PicturePedro Szekely and Craig KnoblockUniversity of Southern California
  4. 4. ProblemSAAMDataWhat ontology to use?Structure mismatchesData consistency What to link to?100% precisionHow to enable museums to do this themselves?Pedro Szekely and Craig KnoblockUniversity of Southern California
  5. 5. Steps to Create Linked Data• Map data to RDF… select ontologies… define mappings• Link to external resources… identify the links• Curate the Linked Data… museums demand 100% correctnessPedro Szekely and Craig KnoblockUniversity of Southern California
  6. 6. select ontologies
  7. 7. University of Southern California
  8. 8. ComplicatedMany irrelevant classesand propertiesIncompleteUniversity of Southern California
  9. 9. edm:ProvidedCHOaac:CulturalHeritageObjectdcterms:creatorore:Aggregationedm:EuropeanaAggregationcrm:E89_Propositional_Objectedm:WebResourceedm:aggregatedCHOedm:hasViewedm:Agent/crm:E39_Actor, foaf:Personaac:PersonrdaGr2:placeOfBirth rdaGr2:placeOfDeathedm:Place/crm:E53_Placeaac:Placeaac:associatedPlaceschema:PostalAddressschema:addressPedro Szekely and Craig KnoblockUniversity of Southern California
  10. 10. edm:ProvidedCHOaac:CulturalHeritageObjectskos:Conceptskos:Conceptedm:hasTypeskos:narrowerskos:prefLabelskos:prefLabelsaam:objectIddcterms:datedcterms:provenancedcterms:rightsdcterms:subjectdcterms:mediumdcterms:titledcterms:descriptiondcterms:creatorore:Aggregationedm:EuropeanaAggregation crm:E89_Propositional_Objectedm:WebResourceedm:aggregatedCHOedm:hasViewedm:Agent/crm:E39_Actor, foaf:Personaac:Personskos:altLabelrdaGr2:dateOfDeathrdaGr2:biographicalInformationrdaGr2:placeOfBirthrdaGr2:placeOfDeathrdaGr2:dateAssociatedWithThePersonedm:Place/crm:E53_Placeaac:Placeaac:associatedPlaceschema:PostalAddressschema:addressCountryschema:addressLocalityschema:addressRegionschema:addressskos:prefLabelschema:Countryschema:namedcterms:formatrdaGr2:dateOfBirthskos:prefLabelsaam:objectNumbersaam:constituentIddcterms:createdPedro Szekely and Craig KnoblockUniversity of Southern California
  11. 11. mapping the data tothe ontologieshow to enable museums to do this themselves?Pedro Szekely and Craig KnoblockUniversity of Southern California
  12. 12. KarmaHierarchicalSourcesServicesModelKarmaTabularSourcesDatabase…Interactive tool for rapidly extracting, cleaning, transforming,integrating, and publishing dataPedro Szekely and Craig KnoblockUniversity of Southern California[ Knoblock, Szekely, et al. Semi-automatically mappingstructured sources into the semantic web. ISWC 2012 ]
  13. 13. specifying transformations andmapping to properties withKarmaPedro Szekely and Craig KnoblockUniversity of Southern California
  14. 14. saam:person/2aac-ont:Person“George M. Aarons”aac-ont:variantNamerdf:typesaam:person/15“Alice Stanley Archeson”aac-ont:marriedNamerdf:typePedro Szekely and Craig KnoblockUniversity of Southern California
  15. 15. Pedro Szekely and Craig KnoblockUniversity of Southern Californiadownload the presentation to view the embedded video
  16. 16. mapping to objectproperties usingKarmaPedro Szekely and Craig KnoblockUniversity of Southern California
  17. 17. Pedro Szekely and Craig KnoblockUniversity of Southern Californiadownload the presentation to view the embedded video
  18. 18. Evaluation of Data Mapping Using KarmaSAAM database8 tables29 columnsOntologies407 classes105 data properties229 object properties# of times Karma’s top 4suggestions contain thecorrect semantic type# of times Karmacorrectly assigns objectpropertiesTime(minutes)Run 1:no trainingdata7 out of 29 (24%) 30 out of 35 (85%) 18Run 2:using Run 1as training27 out of 29 (93%) 32 out of 35 (91%) 8Pedro Szekely and Craig KnoblockUniversity of Southern California
  19. 19. identifying andcurating linksPedro Szekely and Craig KnoblockUniversity of Southern California
  20. 20. Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaMultiple “John Singer Sargent”ima:Person_John_Singer_Sargenta aac-ont:Person ;dct:date "1856-1925" ;foaf:name "John Singer Sargent" .saam:Person_4253a aac-ont:Person ;aac-ont:associatedPlacesaam:SaamPlace_1357324439768t1r13950_0,saam:SaamPlace_1357324439768t1r13951_0 ;saam:constituentId "4253" ;rdaGr2:biographicalInformation“Painter. Sargent traveled …" ;rdaGr2:dateAssociatedWithThePerson "1990-10-1”, "1995-5-8" ;rdaGr2:dateOfBirth "1856-1-12" ;rdaGr2:dateOfDeath "1925-4-15" ;rdaGr2:placeOfBirth saam:SaamPlace_1357324439768t1r13952_0 ;rdaGr2:placeOfDeath saam:SaamPlace_1357324439768t1r13953_0 ;foaf:name "John S. Sargent" ;skos:altLabel "John S. Sargent" ;skos:prefLabel "John Singer Sargent" .cb:Person_John_Singer_Sargenta aac-ont:Person ;ont0:dateOfBirth "1879", "1885" ;ont0:dateOfDeath "1925" ;foaf:name "John Singer Sargent" .met:Person_John_Singer_Sargenta aac-ont:Person ;ont0:placeOfResidence"North and Central America","United States" ;foaf:name "John Singer Sargent" .dallas:Person_John_Singer_Sargenta aac-ont:Person ;ont0:dateOfBirth "1856" ;ont0:dateOfDeath "1925" ;foaf:name "John Singer Sargent" .
  21. 21. Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaJohn Singer Sargentima:SaamPerson_John_Singer_Sargenta saam:SaamPerson ;dct:date "1856-1925" ;foaf:name "John Singer Sargent" .saam:SaamPerson_4253a saam:SaamPerson ;saam:associatedPlacesaam:SaamPlace_1357324439768t1r13950_0,saam:SaamPlace_1357324439768t1r13951_0 ;saam:constituentId "4253" ;rdaGr2:biographicalInformation“Painter. Sargent traveled …" ;rdaGr2:dateAssociatedWithThePerson "1990-10-1”, "1995-5-8" ;rdaGr2:dateOfBirth "1856-1-12" ;rdaGr2:dateOfDeath "1925-4-15" ;rdaGr2:placeOfBirth saam:SaamPlace_1357324439768t1r13952_0 ;rdaGr2:placeOfDeath saam:SaamPlace_1357324439768t1r13953_0 ;skos:altLabel "John S. Sargent" ;skos:prefLabel "John Singer Sargent" .cb:SaamPerson_John_Singer_Sargenta saam:SaamPerson ;ont0:dateOfBirth "1879", "1885" ;ont0:dateOfDeath "1925" ;skos:prefLabel "John Singer Sargent" .met:SaamPerson_John_Singer_Sargenta saam:SaamPerson ;ont0:placeOfResidence"North and Central America","United States" ;foaf:name "John Singer Sargent" .dallas:SaamPerson_John_Singer_Sargenta saam:SaamPerson ;ont0:dateOfBirth "1856" ;ont0:dateOfDeath "1925" ;foaf:name "John Singer Sargent" .
  22. 22. Linking “John Singer Sargent”saam:Person_4253owl:sameAs cb:Person_John_Singer_Sargent ;owl:sameAs dallas:Person_John_Singer_Sargent ;owl:sameAs ima:Person_John_Singer_Sargent ;owl:sameAs met:Person_John_Singer_Sargent ;owl:sameAs dbpedia:John_Singer_Sargent ;owl:sameAs nytimes:N49129220686803623753 ;owl:sameAs w-flick:John_Singer_Sargent ;....Pedro Szekely and Craig KnoblockUniversity of Southern California
  23. 23. IntuitionEstimate discrimination power of properties,e.g., of name, birth and death datesbirth date death date # of people… … …1800 1820 1471800 1821 2841800 1822 213… … …everycombinationof datesSong, D., Heflin, J.: Domain-independent entity coreference for linking ontology instances.ACM Journal of Data and Information Quality (ACM JDIQ) (2012)similar idea toPedro Szekely and Craig KnoblockUniversity of Southern California
  24. 24. Evaluation of Automatic LinkingPedro Szekely and Craig KnoblockUniversity of Southern CaliforniaSAAM names starting with “A” matched by hand 535 people  176 matches
  25. 25. Results of Automatic LinkingGetty ULAN® 2,110Rijksmuseum 551Geonames 3,068DBPedia 2,194New York Times 70Pedro Szekely and Craig KnoblockUniversity of Southern Californiaestimate ≈ 30 missinglinks to DBpedia
  26. 26. Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaCurating Links with Karma
  27. 27. Pedro Szekely and Craig KnoblockUniversity of Southern CaliforniaLinking with Karma
  28. 28. results of automated linking andinteractive curation recorded usingPROVPedro Szekely and Craig KnoblockUniversity of Southern Californiaowl:sameAs statements constructedusing SPARQL CONSTRUCT queriesover PROV records
  29. 29. deploymentPedro Szekely and Craig KnoblockUniversity of Southern California
  30. 30. Pedro Szekely and Craig KnoblockUniversity of Southern California
  31. 31. Pedro Szekely and Craig KnoblockUniversity of Southern California
  32. 32. Pedro Szekely and Craig KnoblockUniversity of Southern California
  33. 33. Pedro Szekely and Craig KnoblockUniversity of Southern California
  34. 34. Related Work• Europeana• 17 million items, 1,500 institutions• Require exports in “Europeana” format• Amsterdam Museum, Museum Finland• Rich ontology, RDF to RDF mapping rules• LODAC museums in Japan• 114 museums, simple ontology• Research Space, British Museum• CIDOC CRM ontologies, complex mappingsWe focused significantly on Linking identification and curation
  35. 35. Next Steps• Applications leveraging linked data• Virtual museum• Tools to create multimedia stories about art• Tools to find inconsistencies• Feed data to wikidata• American Art Collective: a linked dataconsortium of museumsPedro Szekely and Craig KnoblockUniversity of Southern California
  36. 36. Merci

×