Linked Data Activities at OCLC

1
Linked Data Activities at OCLC

• Ralph LeVan
• Senior Research Scientist
• OCLC Research

2
Brief Overview of Linked (Open) Data

• Its about the Links
• Its about the Openness
• Its about the Data
• And there are some infrastructure requirements too

3
Its about the Links

• Books
• http//worldcat.org/oclc/123456
• Classification numbers
• http//dewey.info/class/641/about
• People
• http//viaf.org/viaf/12345679
• Subject headings
• http//tspilot.oclc.org/fast/fst01234567

4
Its about the Openness

• What sort of license is the data available in?

5
Its about the Data

• ltrdfRDFgt
• ltrdfDescription
• rdfabout"http//viaf.org/viaf/12345679"gt
• ltrdftype rdfresource
• "http//xmlns.com/foaf/0.1/Person"/gt
• ltrdftype rdfresource
• "http//RDVocab.info/uri/schema/FRBRentitiesRDA/P
  erson"/gt
• ltfoafnamegtMozziconacci, Jean-Francoislt/foafnam
  egt
• lt/rdfDescriptiongt

6
infrastructure

• Real World Objects
• 303 (See Other) redirects to Generic Objects
• Content negotiation from Generic Objects
• RDF available

7
Dewey

• It should be simple
• http//dewey.info/class/641/
• Browsers get redirected to
• http//dewey.info/class/641/about
• RDF clients get redirected to
• http//dewey.info/class/641/about.rdf

8
But the DDC has a long complex history

• What language did you want?
• http//dewey.info/class/641/about.en
• http//dewey.info/class/641/about.fr
• What edition did you want?
• http//dewey.info/class/641/e22/about
• http//dewey.info/class/641/e23/about
• In what language?
• http//dewey.info/class/641/e22/about.en
• In what format?
• http//dewey.info/class/641/e22/about.en.html
• http//dewey.info/class/641/e22/about.en.rdf

9
Virtual International Authority File VIAF

• It is blissfully simple compared to Dewey!
• http//viaf.org/viaf/12345679
• Generates a 303 redirect to
• http//viaf.org/viaf/12345679/
• Where content negotiation will get you either
• http//viaf.org/viaf/12345679/viaf.html
• http//viaf.org/viaf/12345679/viaf.xml
• http//viaf.org/viaf/12345679/viaf.rss
• http//viaf.org/viaf/12345679/viaf.rdf
• http//viaf.org/viaf/12345679/marc21.xml
• http//viaf.org/viaf/12345679/unimarc.xml

10
Elegant (but incomplete) Data

• ltrdfRDFgt
• ltrdfDescription
• rdfabout"http//viaf.org/viaf/12345679"gt
• ltrdftype rdfresource
• "http//xmlns.com/foaf/0.1/Person"/gt
• ltrdftype rdfresource
• "http//RDVocab.info/uri/schema/FRBRentitiesRDA/P
  erson"/gt
• ltfoafnamegtMozziconacci, Jean-Francoislt/foafnam
  egt
• lt/rdfDescriptiongt

11

• ltskosConcept rdfabout"http//viaf.org/viaf/
  sourceID/BNF7C12153518skosConcept"gt
• ltskosinScheme rdfresource"http//viaf.org/
  authorityScheme/BNF"/gt
• ltskosprefLabelgtMozziconacci, Jean-Francois
• lt/skosprefLabelgt
• ltfoaffocus rdfresource"http//viaf.org/viaf/12
  345679"/gt
• lt/skosConceptgt

12
FAST Faceted Application of Subject
Terminology, subject headings

• Another elegant URI pattern
• http//tspilot.oclc.org/fast/fst01234567
• http//tspilot.oclc.org/fast/fst01234567.j
  son
• http//tspilot.oclc.org/fast/fst01234567.m
  ads
• http//tspilot.oclc.org/fast/fst01234567.m
  arcxml
• http//tspilot.oclc.org/fast/fst01234567.s
  kos
• http//tspilot.oclc.org/fast/fst01234567.z
  thes

13
But, a work in progress

• No RDF (but existing SKOS shows it wont be hard)
• Camped on the tspilot.oclc.org domain

14
Not WorldCat Identities

• http//worldcat.org/identities/lccn-n79-91588
• Why not?
• Limited product development resources
• No customer demand
• What do you mean no customer demand?!?!
• Our library customers prefer MARC or simple XML
• instead of RDF
• The potential consumers who want Identities as
• linked data expect to get it for free

15
W3C Library Linked Data Incubator Group

• Andy Houghton
• Tod Matola
• Michael Panzer
• Jeff Young
• The mission of the group is to help increase
  global interoperability of library data on the
  Web, by bringing together people involved in
  Semantic Web activitiesfocusing on Linked
  Datain the library community and beyond
• http//www.w3.org/2005/Incubator/lld/

16
Open Source Infrastructure

• Our VIAF interface technology is available as
  Open Source at http//code.google.com/p/oclcsrw/
• It provides the ability to expose records stored
  in Lucene (or a similar database) as Linked Data
• Other technologies (such as D2R) are available
  for data in relational databases

17
Turning Content into Links

• If Linked Data is about links, then how do you
  turn your content into links?
• Google Refine and the ReconciliationServiceAPI
• Refine takes spreadsheets of data and
• supplements that data with links
• It uses the ReconciliationServiceAPI to query a
• database and get back links
• Think of OpenURL for the Linked Data world

18
Poised for success

• OCLC has the data, the infrastructure and the
  knowledge to participate in the Linked Data
  world.