Hyperautomation and AI/ML: A Strategy for Digital Transformation Success.pdf
ORCID Outreach meeting Oxford may 2013 integration demo
1. orcid.org
Integration demo
ODIN: Linking ORCID and DataCite
Gudmundur A.Thorisson, PhD
ORCID EU / ODIN Project
and University of Iceland Computing Services
g.thorisson@orcid-eu.org
http://gthorisson.name | http://orcid.org/0000-0001-5635-1860
This work is published under the Creative Commons Attribution
license (CC BY: http://creativecommons.org/licenses/by/3.0/)
which means that it can be freely copied, redistributed and adapted,
as long as proper attribution is given.
2. ODIN and ORCID EU
• ORCID and DataCite Interoperability
Network
• Aim: connect researchers and datasets via
persistent identifiers
• 2 year EC-funded project
• 7 partners (5 EU + 2 international)
• ORCID EU incorporated in Belgium
http://odin-project.eu
Odinn the Wanderer
Credit: Wikipedia Commons
3. Technical development in ODIN
• ODIN ≠ infrastructure creation project! Output
is roadmaps, technical reports, prototypes etc.
• Proof-of-principle implementations
• British Library: use and citation of social sciences
data, long-running British Birth Cohort Studies
• CERN: attribution in high-energy physics - dealing
with large no. of authors on HEP papers & datasets
• ORCID EU “Labs” mini-projects
• Work Package 4 - Interoperability
4. .. three main threats or “items of unfinished business” emanating
from lack of recognition of the need for robust ways of
identifying contributors and their data in e-Science
• Inability to follow interconnections between datasets and
contributors as a method of data discovery.
• Inability to share and connect identifiers of contributors and
authors between different user communities.
• Inability to uniquely identify datasets attributed to a
particular contributor and contributors to a particular dataset.
Linking contributors and datasets
5. Linking contributors and datasets
Embedding ORCiDs in
data centre workflows
Prospective - long
term aim
6. Linking contributors and datasets
Embedding ORCiDs in
data centre workflows
Prospective - long
term aim
Ad hoc claiming of already
published datasets
Retrospective - necessary
now and for quite some time
7. Terminology
• “claiming” = asserting that one has authored or
otherwise contributed to a published work
• “I authored this paper or book”
• “I created this dataset”
• “I had role X in the generation of this dataset and
contributedY and Z”
• Focus here on self-claims - claims by
contributors themselves about their works
8. Doing it
• Existing resources available to build on
• CrossRef Metadata Search tool http://search.crossref.org
• DataCite search API
• ORCID Member API
• Technical approach
• Create simple standalone, loosely coupled tool
• Connect CrossRef MDS to DataCite search API
• Add several functionality and UI tweaks
• Aim: to get work metadata into user’s ORCID
profile
10. Challenges / obstacles
• Metadata quality and consistency
• DataCite vs. CrossRef vs. ORCID
• Basic things like date and author is handled in subtly
different ways in DOI metadata
• Work types in ORCID hardcoded to BibTeX - w00t!
• ORCID Works Metadata Working Group
discussing these and related issues
• how can ORCID best align with and reuse/adapt
existing standards & community ontologies/
vocabularies?
11. Not yet another standard please!
Credit: http://xkcd.com/927/
12. The open source advantage
• Key building blocks
• Generic - Sinatra webapp framework, OmniAuth
authentication framework [~dozen more]
• Domain-specific - CrossRef Metadata Search,
omniauth-orcid Ruby gem
• Reuse/Remix/Extend to build new things
• ORCID to take advantage of open source
• https://github.com/ORCID/ORCID-Source
• http://orcid.org/about/community/orcid-technical-community
13. Acknowledgements
• Martin (the other half of ORCID
EU Labs, did majority of work!)
• Karl & Geoff @CrossRef for open-sourcing
• ODIN partners for testing&feedback
• Numerous other testers in the community
14. Published under the CC BY license (http://
creativecommons.org/licenses/by/3.0/)
