Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Introduction to the Linked Art Data Model

464 views

Published on

An introduction to the linked.art LOD data model, based on a carefully selected profile of CIDOC-CRM, and expressed as JSON-LD. It focuses on developer happiness and data usability, while trying to also maintain as much of the richness of CRM as possible.

Published in: Technology
  • Be the first to comment

Introduction to the Linked Art Data Model

  1. 1. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Introduction  to  the Linked  Art  Data  Model Rob  Sanderson,  David  Newbury Semantic  Architect,  Software  &  Data  Architect J.  Paul  Getty  Trust rsanderson@getty.edu,  dnewbury@getty.edu @azaroth42,  @workergnome
  2. 2. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu CIDOC-­‐CRM,  RDF,   &  Linked  Data
  3. 3. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu What  is  the  CIDOC-­‐CRM? “The  CIDOC-­‐CRM  provides  definitions  and  a  formal   structure  for  describing  the  implicit  and  explicit  concepts   and  relationships  used  in  cultural  heritage  documentation.” • An  event-­‐based way  to  look  at  the  world • Standardized  meaning  and  structures  for  information • ”Semantic  glue"  between  different  information  sources • Highly  theoretical
  4. 4. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu What  is  RDF? “RDF  is  a  standard  model  for  data  interchange  on  the  Web.” • A  way  for  computers  to  work  with  facts • A  way  to  express  statements  about  resources • A  W3C  (web-­‐based,  web-­‐friendly)  data  standard   Subject -­‐>  Predicate -­‐>  Object
  5. 5. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu What  is  Linked  Data? “The  term  Linked  Data  refers  to  a  set  of  best  practices  for   publishing  structured  data  on  the  Web.” • Some  rules  that  help  make  RDF  useful  for  others • Every  ID  is  a  HTTP  URL,  and  it  returns  info  about  itself • You  should  use  other  people’s  URLs  whenever  you  can Subject (URI) -­‐>  Predicate (URI) -­‐>  Object  (URI)
  6. 6. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu What  Came  Before? CIDOC-­‐CRM  Linked  Data  models  by  the  British  Museum,   YCBA,  SAAM,  Getty,  Zeri,  and  others. Important  precursors  and  references  for  our  model, these  were: • Publishing-­‐focused • Institution-­‐specific models  and  predicates • Focused  on  Semantic  Correctness
  7. 7. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Linked  Art  Model A  CIDOC-­‐CRM  Linked  Data  model  designed  to  work  across   many  museums  and  to  enable  functional  applications  to  be   built  on  top  of  the  model. • Standardized model  across  many  institutions • Supports  varying  levels  of  completeness • Enables  “round-­‐trippability”  of  data • Uses  CIDOC  alongside  AAT  and  other  RDF  vocabularies • Focused  on  Usability
  8. 8. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Linked  Art  Model Linked  Art  provides  patterns  and  models  that  enable   cultural  heritage  institutions  to  easily  publish  their  data  for   both  event-­‐based  digital  research  projects  and  for  use  by   non-­‐cultural-­‐heritage  developers. 90%  of  the  use  cases  of  90%  of  the  organizations,   with  only  10%  of  the  complexity  of  the  full  CRM  ontology.
  9. 9. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Patrick  Hochstenbach,  @hochstenbach
  10. 10. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Usable vs  Complete
  11. 11. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Target  Zone
  12. 12. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Successful  Data  Models… • Solve  actual  challenges,  documented  as  use  cases • Using  data  that  is  captured  and  available • Allow  consistent  description  of  shared  use  cases • Allow  for  addition  of  information  not  shared • Can  be  productively  used   • Via  easy-­‐to-­‐implement  services • With  easy-­‐to-­‐implement  applications • Provide  interoperability with  other  data
  13. 13. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Successful  Data  Models… Are  developed  … • Iteratively • We  will  not  get  it  right  first  time • Responsively • We  will  adapt  it  in  response  to  feedback • Responsibly • We  will  consider  changes/features  carefully • Collaboratively • We  will  engage  with  the  community  and  stakeholders
  14. 14. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Successful  Data  Models… Have  carefully  bounded  scope Out  of  scope: • Complex  Bibliography • Adopt  eventual  solution  from  Library  domain • Data  Provenance • Not  Museum  specific,  adopt  solution  from  LOD • Quantification  of  Uncertainty • Unsustainably  complex,  data  unavailable
  15. 15. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Successful  Data  Models… Focus on  the  primary  resources  of  concern  for  the  domain Core  Classes: • Object • Activity • Person
  16. 16. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Successful  Data  Models… Are  Consistent and  Unsurprising Common  Patterns  designed  and  used: • Classification  of  Specifics • Partitioning • Statements  about  Resources
  17. 17. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Object
  18. 18. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Person
  19. 19. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Activity
  20. 20. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Classification  of  Specifics We  use  AAT  to  provide  more  specific  classifications  of   resources  than  are  possible  using  CIDOC-­‐CRM • Object:  Painting  vs  Sculpture • Identifier:  Accession  Number  vs  Database  Id • Place:  Country  vs  City • Organization:  Museum  vs  Art  Dealer • Activity:  Engraving  vs  Casting
  21. 21. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Partitioning Many  types  of  resource  can  be  partitioned,  creating   identity  for  further  description  of  more  granular  parts • Objects:  Parts,  Features  with  different  dimensions  etc • Activities:  Exhibition  vs  Venue,  Auction  vs  Lot • Locations:  Country,  Region,  City,  District,  Building • Texts:  Set,  Volume,  Chapter,  Page • Organizations:    Organization,  Program,  Dept,  Group
  22. 22. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Statements  about  Resources Resources  can  have  statements  (notes)  with  classifications   for  the  topic  of  the  statement.  Can  co-­‐exist  with  machine   readable  data.   • Materials • Provenance • Attribution • Biography • Description
  23. 23. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu JSON-­‐LD  Representation { "@context": "https://lod.museum/ns/context/1/full.jsonld", "id": "https://lod.museum/example/object/1", "type": "ManMadeObject", "classified_as": "aat:300033618", # by reference "label": "Example Painting", "made_of": { "id": "aat:300015045", # by (minimal) value "type": "Material", "label": "watercolor" } }
  24. 24. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Linked.Art site Detailed  information  about  the  model  is  on  the  site: http://linked.art/model/ Split  into  sections  based  on  the  objects  of  interest: • Objects;  Collections • People;  Places • Provenance;  Exhibitions • Sources;  Datasets
  25. 25. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Linked.Art vs  AAC  Target  Model? The  linked.art model  continues  the  work  of  the  AAC  target   model: • Apply  lessons  learnt  from  use  and  application  to   different  datasets,  domains  (Provenance,  Photos) • Provides  more  detail  based  on  questions  from  reviews   and  subsequent  AAC  partner  discussion • Ongoing  support  and  engagement  with  the  community   beyond  AAC  – Provenance,  Pharos,  Nomisma,  ...
  26. 26. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Thank  You! Rob  Sanderson,  David  Newbury rsanderson@getty.edu,  dnewbury@getty.edu @azaroth42,  @workergnome
  27. 27. @azaroth42 rsanderson @getty.edu IIIF:  Interoperabilituy Intro  to  linked.art Data  Model   @azaroth42 rsanderson @getty.edu Discuss!

×