GLAM Rocks! London Semantic Web Meetup
Download as PPTX, PDF•3 likes•5,076 views
This document summarizes a meetup event on semantic web and linked data for cultural heritage institutions in London on March 28, 2012. It lists presentations on projects linking archival and museum collections using semantic web technologies, including LOCAH, Linking Lives, and the British Library's and British Museum's linked data efforts. Challenges discussed include identity resolution, data modeling complexity, sustainability, and licensing. The final presentation was given by Adrian Stevenson of the University of Manchester on bibliographic frameworks.
GLAM Rocks! London Semantic Web Meetup
- 1. GLAM ROCKS UK 2012 london semantic web meetup bbc, white city, london, uk 28th march 2012 #londonsemweb adrian stevenson senior technical innovations coordinator mimas, university of manchester, uk @adrianstevenson
- 9. BRITISH LIBRARY www.bl.uk/bibliographic/datafree.html #sxsw #lodlam
- 10. BRITISH MUSEUM collection.britishmuseum.org #sxsw #lodlam
- 11. DISCOVER Y discovery.ac.uk
- 12. BRICOLAG E bricolage.ilrt.bris.ac.uk #sxsw #lodlam
- 13. MORE GLAM ROCK • archon – kasabi.com/dataset/archon/ • SALDA – blogs.sussex.ac.uk/salda/ • openArt – yorkdl.wordpress.com/category/openart • OEM-UK - ioeoemuk.wordpress.com/ • contextual wrappers - contextualwrappers2.wordpress.com/ • openbib – openbiblio.net/ • comet – cul-comet.blogspot.co.uk/ • trenches to triples – trenchestotriples.blogspot.co.uk/ • open Book – jiscopenbook.wordpress.com/about • digital.bodlean
- 14. DISCOVERY WW1 EXEMPLAR jiscww1.jiscinvolve.org/wp/jisc-ww1-discovery-programme #sxsw #lodlam
- 15. TRICKY STUFF
- 16. WHERE’S THE DATA? “you shouldn’t be allowed to publish linked data unless you’re using some” chris gutteridge, university of southampton
- 17. NAME MATCHING <reg dwight> <sameAs> <??> <james osterberg> <sameAs> <??> <j a genzale> <sameAs> <??> <stuart goddard> <sameAs> <??>
- 18. MORE TRICKY STUFF • dirty data, URI persistence. • data modelling can be hard – Steep learning curve – Complexity • how sustainable are the data sources? • how scalable are triple stores? • can you track the provenance of data sources? • are data licensing issues covered?
- 19. BIBLIOGRAPHIC FRAMEWORK www.loc.gov/marc/transition/news/framework-103111.html
- 20. adrian stevenson mimas, university of manchester, uk [email protected] www.mimas.ac.uk @adrianstevenson www.linkedin.com/in/adrianstevenson www.slideshare.net/adrianstevenson
- 21. attribution and cc license sections of this presentation adapted from materials created by other members of the LOCAH & Linking Lives projects this presentation available under creative commons non commercial-share alike http://creativecommons.org/licenses/by-nc/2.0/uk/
Editor's Notes
- #2: Mention GLAM acronymIntroduce myself, Mimas, & JISCSay Mimascentre for innovations and services in the data spaceServices include archives and lib, also several census services, incl census dissem unit, economic data from world bank , ESDS geo-spatial dataWeb of KnowledgeUK Pub Med CentralVery interested in ways of doing data better – hence LD and sem web interestHave restricted myself to RDF based Linked Data for this meeting
- #3: Putting out LD for Mimas Archives Hub servicedata about national archival holdingsGround breaking LD project that influenced a number of folowing projectsData dump, SPARQLCreative Commons zeroWe are currently enhancing and expanding the datasetBlog has really useful pragmatic stuff on how to do linked data. A number of pplincl the BL people have referred to as their bible.
- #4: LOCAH continuation projectBased around namesAims to show the value of consuming Linked DataBrings datasets together around names
- #5: Lots of good biography info contained in the archives that are somewhat hidden and we want to pull out
- #7: Say what VIAF is – interested in pulling these images in, but unsure what we can pull in. easier if things explicity openWhy Linking Lives?Giving (more) control to others to use the data. Not locked in our interface.Placing archives in a global information spaceExternal data enriches our dataDynamic links to other contentExtensibleAn exemplar – shows what can be done
- #8: ‘Every story has a beginning’Nice example of consumption of Archives Hub linked data
- #9: Meet to see whjere things at with LOD ArchivesMain outcome was design patterns.First meet under auspice of CDOC CRMAttempt to start to do things in a similar way in LAM areas. Take a look discreter parts of data models in LAM area and see what have in common, to help people starting up. E.g. ‘held at’“explore the questions that are asked by cultural heritage researchers”“design patterns for fragments of the Linked Data graph”
- #11: Mainly SPARQL interface.It provides access to the same collection data available through the Museum’s web presented Collection Online, but in a computer readable format.The data has also been organised using the CIDOC-CRM (Conceptual Reference Model)
- #12: A national initiative, funded by JISCIncludes major research universities, The National Archives, British Library & Collections Trust….Libraries, Archives & Museums spaceHas been funding various actriviies and projectsIt’s about principles of open data and aggregatable open data. Not specifically linked data RDF, but open to those principles and some projects are RDF as will outline here
- #13: publish catalogue metadata as Linked Open Data:The Penguin Archive, a comprehensive collection of the publisher’s papers and books.The Geology Museum, a 100,000 specimen collection housing many unique and irreplaceable resources.main outputs of this project:Linked Data sets for the Geology Museum and the Penguin Archive, with reuse guidance.Two demonstrators illustrating data reuse: a browser-based mapping application for exploring the Geology collection via its geography, and an interactive timeline displaying the chronology of selected resources within the Penguin Archive.
- #17: Gutteridge “You souldn’t be allowed to publish linked data unless you’re using some”
- #18: The old co-referenceproblmeCan be very tricky and time consumingCrowdsourcing a possibility
- #19: Steep learning curve: - RDF Linked Data modelingterminology - Lack of archive domain examples – though you now have LOCAH! - Certain level of expertise neededDirty Data - Joe Bloggs and others’ rather than just a name, or where the access points do not have rules or a source associated with them. - Extent data highly variableComplexity - “lower level” units interpreted in context of the higher levels of description - Arguably “incomplete” without the contextual data.Relations are asserted, e.g. member-of/component-ofBut there is no requirement or expectation that data consumers will follow the links describing the relations
- #21: Mention GLAM acronymIntroduce myself, Mimas, & JISCSay Mimascentre for innovations and services in the data spaceServices include archives and lib, also several census services, incl census dissem unit, economic data from world bank , ESDS geo-spatial dataWeb of KnowledgeUK Pub Med CentralVery interested in ways of doing data better – hence LD and sem web interestHave restricted myself to RDF based Linked Data for this meeting
- #22: Mention GLAM acronymIntroduce myself, Mimas, & JISCSay Mimascentre for innovations and services in the data spaceServices include archives and lib, also several census services, incl census dissem unit, economic data from world bank , ESDS geo-spatial dataWeb of KnowledgeUK Pub Med CentralVery interested in ways of doing data better – hence LD and sem web interestHave restricted myself to RDF based Linked Data for this meeting