Linked Data for Libraries: Benefits of a Conceptual Shift from Library-Specific Record Structures to RDF-based Data Models (IFLA-2012)


Published on

This presentation (full text paper: ) provides recommendations for making a conceptual shift from current document-centric to data-centric metadata. The importance of adjusting current library models such as Resource Description and Access (RDA) and Functional Requirements for Bibliographic Records (FRBR) to models based on Linked Data principles is discussed. In relation to technical formats, the paper suggests the need to leapfrog from Machine Readable Cataloguing (MARC) to Resource Description Framework (RDF), without disrupting current library metadata operations.

Published in: Technology, Education
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Linked Data for Libraries: Benefits of a Conceptual Shift from Library-Specific Record Structures to RDF-based Data Models (IFLA-2012)

  1. 1. Getaneh Alemu, Brett Stevens, Penny Ross and Jane ChandlerWorld Library and Information University of Portsmouth, Portsmouth, UKCongress 78th IFLA GeneralConference and Assembly11-17 August 2012, Helsinki,Finland
  2. 2. Library of Ashurbanipal Library of Alexandria(Blair, 2010; Denton, 2007; Dunsire, 2009; IFLA, 2009; Lubetzky, 1953; Morville , 2005; Svenonius, 2000; Wright, 2007)
  3. 3. (IFLA, 2009; Svenonius, 2000)
  4. 4.  Record centric, more attuned to human consumption rather than machine processing  Metadata duplication  Data inconsistency  Lack of granularity  Identification, naming, terminological issues  Lack of scalability to the web(Coyle, 2010; Coyle & Hillmann, 2007; Lagoze, 2010; Mathes, 2004; Shirky, 2005; Veltman, 2001; W3C, 2011; Weinberger, 2005, 2007)
  5. 5. MARC MARC must die! (Tennant, 2002) MARC is deeply embedded in library systems and functions; Making any changes too difficult and expensive; Alternative formats, including XML fail to deliver the additional functionality required to merit and justify the changeover; MARC persisted but it is considered to be inadequate and anachronistic. (Coyle, 2010; Coyle & Hillmann, 2007; LC, 2011; Wallis, 2011a, 2011b)
  6. 6.  Deriving lighter schema from a complex one (Chan & Zeng, 2006)
  7. 7.  Incorporating diversity/multiple view points  “Cultural diversity is as vital as biodiversity” (Veltman, 2001) Open-world assumption  “Anyone can say anything about any topic” resulting in “variations and disagreements” (Allemnag & Hendler, 2008)  Distributed web of data  The network effect  Linked Data principles (Berners-Lee, 2009)  Global unique identification  Use of scalable data model  Use of consistent technical formats  Scalable in-bound and out-bound linking
  8. 8.  RDF  RDFS  OWL  SPARQL  URI(Allemnag & Hendler, 2008; Berners-Lee, 1997; W3C, 2004a, 2004b)
  9. 9.
  10. 10.  Silos: library-domain-specific languages and terminologies Conflation between metadata content and metadata presentation/display Technical complexity of Linked Data technologies such as RDF/XML, RDFS, OWL and SPARQL Generation, maintenance, resolution and preservation of URIs and namespaces
  11. 11.  Linked Data is about metadata and libraries have always been creating and managing metadata Libraries are interested in Linked Data  The Europeana Data Model  British Library Data Model  Virtual International Authority File (VIAF)  Library of Congress Subject Headings (LCSH)  Open Metadata Registry  Lexvo (URI referenced controlled list of characters, words, terms)  GeoNames (geographical database)  MARC country and language codes  (top level classes of Dewey Decimal Classification)  RDF book mash-up (information about books and their authors) (The British Library, 2011; Wallis, 2011)
  12. 12.  Metadata openness and sharing Facilitate serendipitous discovery of information resources Identification of resource usage patterns, zeitgeist and emergent metadata Facet-based navigation Metadata enriched with links
  13. 13. Source: OReilly (2005) Reusable, re-mixable/mashable metadata Application profiles Linking to non-library information sources
  14. 14. (Anhalt & Stewart, 2011; Coyle & Hillmann, 2007; Dunsire, 2009; W3C, 2004)
  15. 15.  Ontologies  Centralised vs decentralised  Complete vs good enough  Focus on describing entities  Develop vocabularies  Properties
  16. 16. The richer an object is described with metadata, the more likely that it conforms to the multitude of perspectives and interpretations of users. (Alemu, Stevens, & Ross, 2012) Otlet’s “The social space of documents” “Réseau: a tool to create semantic links between documents and keep track of the annotations made by readers, eventually forming new trails of documents, which he calls “the book about the book” (Wright, 2007)