際際滷

際際滷Share a Scribd company logo
ANDS: increasing
connections to add value
Andrew Treloar, Director of Technology
1
ANDS at a glance
 In operation since 2009
 40 staff (mostly Melbourne, Canberra)
 Funded as the voice for data
 Working to make Australias research data more
valuable
 Funding through to mid 2016 (mid 2017?)
CC-BY @atreloar 2
ANDS enables transformation of:
Data:
Unmanaged
Disconnected
Invisible
Single use
CC-BY @atreloar 3
to Structured Collections:
Managed
Connected
Findable
Reusable
so that Australian researchers can easily publish,
discover, access and use/re-use research data.
CC-BY @atreloar
4
Why care about connections?
 Context
 Discovery
 Evidence for trustworthiness
 Credit
 Integrity of scholarly record
 Better basis for analysis
 Summary: greater value!
5
Central role of identifiers
 Need to be able to identify the endpoint for a
connection
 ideally uniquely
 Some identifiers need to be managed globally, but
some can be managed locally
 NOTE: Identifiers do not necessarily need to
persist
6
5 things that need to persist for persistence
 the identifier- that is the token or string itself.
 the thing being pointed at by the identifier. For example,
the content.
 the mapping of the identifier to the thing being identified.
 the resolver that allows one to follow the mapping of the
identifier to the thing being identified.
 a mechanism for updating the mapping of the identifier to
the thing being identified.
 See http://crosstech.crossref.org/2015/03/january-2015-
doi-outage-followup-report.html for more on persistence
7
8
vs
Persistence vs. Coolness
 Persistent identifiers are often criticised in favour
of cool URIs
 http://www.w3.org/TR/cooluris/
 Yes, its possible to be cool, but impossible to be
effortlessly cool
 Persistent identifier mechanisms are often a way
of engineering away brittleness in identifier
infrastructure
9
A data-centric view of the space
10
ResearcherLink
11
 ANDS originally encouraged use of Trove/NLA ID
 ORCID now has market momentum behind it
 ANDS recently assisted with work towards decision
by ARC/NHMRC/UA/CAUL/ARMS to broadly
encourage use of ORCID
 doi:10.4225/14/552b16efae74b
PublicationLink
12
 DataCite DOIs increasingly the way to link from
publications to data
 ANDS working with Thomson-Reuters Data
Citation Index to enable them to track citations to
data
 ANDS provides national harvest point through
Research Data Australia
 http://ands.org.au/cite-data/dci.html
GrantLink
13
 ANDS working with ARC and NHMRC to provide
LOD URIs for research grants
 ARC: http://purl.org/au-research/grants/arc/DP0211416
 NHMRC: http://purl.org/au-
research/grants/nhmrc/1020281
 Overview: http://ands.org.au/guides/ardc-activity-
infrastructure.html
 API:
http://developers.ands.org.au/services/getgrants/
Research Data Switchboard
14
 Collaborative project to enable cross-platform
discovery between research data services
 Created as part of Research Data Alliance DDRI-
WG
 Using network of AMS EC2 instances to automate
process
 Harvest from data providers, create graph
database, provide API to consumers of graph
Obligatory architecture diagram
15
Some gratuitous eye-candy
16
17
Project THOR
Technical and Human infrastructure for Open
Research
Builds on ODIN (ORCID-DataCite Interoperability
Network)
Establishing seamless integration between articles,
data, and researchers across the research lifecycle
Making persistent identifier use for people and
research artefacts the default
http://project-thor.eu
Project partners
Project Partners
Conclusion
20
 ANDS cares about connections
 ANDS is working to develop and provide LOD-
compatible URIs as part of national infrastructure
 ANDS is continuing development on Research Data
Switchboard
 available for public use on September 23
Questions?
 ands.org.au
 http://ands.org.au/guides/data-connections.html
 http://rd-switchboard.org/
 andrew.treloar@ands.org.au
 @atreloar
21

More Related Content

ANDS: Increasing connections to add value

  • 1. ANDS: increasing connections to add value Andrew Treloar, Director of Technology 1
  • 2. ANDS at a glance In operation since 2009 40 staff (mostly Melbourne, Canberra) Funded as the voice for data Working to make Australias research data more valuable Funding through to mid 2016 (mid 2017?) CC-BY @atreloar 2
  • 3. ANDS enables transformation of: Data: Unmanaged Disconnected Invisible Single use CC-BY @atreloar 3 to Structured Collections: Managed Connected Findable Reusable so that Australian researchers can easily publish, discover, access and use/re-use research data.
  • 5. Why care about connections? Context Discovery Evidence for trustworthiness Credit Integrity of scholarly record Better basis for analysis Summary: greater value! 5
  • 6. Central role of identifiers Need to be able to identify the endpoint for a connection ideally uniquely Some identifiers need to be managed globally, but some can be managed locally NOTE: Identifiers do not necessarily need to persist 6
  • 7. 5 things that need to persist for persistence the identifier- that is the token or string itself. the thing being pointed at by the identifier. For example, the content. the mapping of the identifier to the thing being identified. the resolver that allows one to follow the mapping of the identifier to the thing being identified. a mechanism for updating the mapping of the identifier to the thing being identified. See http://crosstech.crossref.org/2015/03/january-2015- doi-outage-followup-report.html for more on persistence 7
  • 9. Persistence vs. Coolness Persistent identifiers are often criticised in favour of cool URIs http://www.w3.org/TR/cooluris/ Yes, its possible to be cool, but impossible to be effortlessly cool Persistent identifier mechanisms are often a way of engineering away brittleness in identifier infrastructure 9
  • 10. A data-centric view of the space 10
  • 11. ResearcherLink 11 ANDS originally encouraged use of Trove/NLA ID ORCID now has market momentum behind it ANDS recently assisted with work towards decision by ARC/NHMRC/UA/CAUL/ARMS to broadly encourage use of ORCID doi:10.4225/14/552b16efae74b
  • 12. PublicationLink 12 DataCite DOIs increasingly the way to link from publications to data ANDS working with Thomson-Reuters Data Citation Index to enable them to track citations to data ANDS provides national harvest point through Research Data Australia http://ands.org.au/cite-data/dci.html
  • 13. GrantLink 13 ANDS working with ARC and NHMRC to provide LOD URIs for research grants ARC: http://purl.org/au-research/grants/arc/DP0211416 NHMRC: http://purl.org/au- research/grants/nhmrc/1020281 Overview: http://ands.org.au/guides/ardc-activity- infrastructure.html API: http://developers.ands.org.au/services/getgrants/
  • 14. Research Data Switchboard 14 Collaborative project to enable cross-platform discovery between research data services Created as part of Research Data Alliance DDRI- WG Using network of AMS EC2 instances to automate process Harvest from data providers, create graph database, provide API to consumers of graph
  • 17. 17
  • 18. Project THOR Technical and Human infrastructure for Open Research Builds on ODIN (ORCID-DataCite Interoperability Network) Establishing seamless integration between articles, data, and researchers across the research lifecycle Making persistent identifier use for people and research artefacts the default http://project-thor.eu
  • 20. Conclusion 20 ANDS cares about connections ANDS is working to develop and provide LOD- compatible URIs as part of national infrastructure ANDS is continuing development on Research Data Switchboard available for public use on September 23
  • 21. Questions? ands.org.au http://ands.org.au/guides/data-connections.html http://rd-switchboard.org/ andrew.treloar@ands.org.au @atreloar 21

Editor's Notes

  • #4: So, let me now discuss institutional responses to all of this in the context of the ANDS Four Transformations
  • #11: and we could add models/software/workflows to this diagram as well if we wanted
  • #15: What kind of identifiers do we support? DOI ORCID ScopusID PURL (grants) In addition we use URLs from Research Data Australia, for Australian collections with no DOI. Does ANDS has any plan to include ISNI in this project? We had such a plan, but at this stage our funding is limited. We have to focus on the core functionality and the components that are directly connects the switchboard to our DDRI partners (DRYAD, CERN, da-ra, DCU, figshare and DataCite) until September P6. After that we can incorporate more new developments. When will this system be available for public use? September 23