1. EOSC-Nordic project has received funding from the European Union’s Horizon 2020
research and innovation programme under grant agreement No 857652
FAIR support for Data
Repositories
Josefine Nordling, WP4 FAIR Data
Presentation slides available:
https://www.fairdata.fi/koulutus/koulutuksen-tallenteet/
2. By Taller345 - Own work, CC
BY-SA 4.0,
https://commons.wikimedia
.org/w/index.php?curid=63
245736
Iceland
• UNIVERSITY
OF ICELAND
Norway
• NORDFORSK
• UNINETT SIGMA2 AS
• NORWEGIAN CENTER
FOR RESEARCH DATA
Denmark
• DENMARK TECHNICAL
UNIVERSITY
• UNIVERSITY OF SOUTHERN
DENMARK
• DANISH NATIONAL ARCHIVES
• UNIVERSITY OF COPENHAGEN
• CAPITAL REGION OF
DENMARK
• NORDUNET / AS
Finland
• CSC – IT CENTER OF
SCIENCE
• UNIVERSITY OF HELSINKI
• UNIVERSITY OF TAMPERE
• UNIVERSITY OF EASTERN
FINLAND
• FINNISH
METEOROLOGICAL
INSTITUTE
Sweden
• UNIVERSITY OF UPPSALA
• SWEDISH RESEARCH
COUNCIL
• UNIVERSITY OF
GOTHENBORG
Estonia
• UNIVERSITY OF TARTU
• NATIONAL INSTITUTE OF
CHEMICAL PHYSICS AND
BIOPHYSICS
Latvia
• RIGA TECHNICAL
UNIVERSITY
Lithuania
• UNIVERSITY OF VILNIUS
Netherlands
GoFair
Germany
DKRZ
24 Participants
3. 1
2
3
4
5
Support coordination, harmoni-
sation and alignment of Nordic
and Baltic national policies and
practices related to the provision
of horizontal research data
services with EOSC
Increase the discoverability of Nordic
& Baltic services. Extend and expand
their use by making them accessible
through the EOSC portal
MAIN
OBJECTIVES
Promote and support the
uptake of FAIR data practices
and certification schemas
across the Nordics
Accelerate the progress and
attractiveness of EOSC by piloting
& delivering innovative solutions
developed and tested in a useful
and functional cross-border
environment
Provide a Knowledge Hub
to deliver training and technical
support to new service providers
and communities willing to
engage with EOSC during and
after the project lifetime
OBJECTIVE 1
OBJECTIVE 5 OBJECTIVE 4
OBJECTIVE 3
OBJECTIVE 2EOSC NORDIC
4. Implementing FAIR in the research
community – How?
• One of the missions of EOSC-Nordic is to ‘implement FAIR’ in the
region. This implementation will happen primarily by:
1. Disseminating the benefits of going FAIR to a broad science
community,
2. Providing advice on how to concretely address the FAIRification
challenge and
3. Supporting the communities by hosting a handful of hackathons
and/or metadata4machines events.
5. Implementing FAIR in the research
community – Approaches
• The FAIR support measures:
1. Machine-readable FAIR maturity evaluations of data repositories
2. FAIR data standards and (FAIR) certification schema
3. Incentives for uptake of FAIR practices
6. Implementing FAIR in the research
community – FAIR support for services
• The FAIR support measures:
1. Machine-readable FAIR maturity evaluations of data repositories
2. FAIR data standards and (FAIR) certification schema
3. Incentives for uptake of FAIR practices
7. FAIR maturity evaluations – selection
criteria & requirements
• Based on re3data.org and associated with a Nordic & Baltic member state
• Additionally, an internal survey among the 14 partners, listing repositories
that host research data of any kind (raw, processed, structured or
unstructured)
• We excluded repositories that only host publications (typically pre-prints or
reprints) and those that have no clear potential use in science
• This yielded 134 repositories associated with one or more of the Nordic +
Baltic countries. The repositories can be broad cross-discipline repositories,
larger deposits for infrastructures or community specific repositories that
are limited to a narrow scientific field
• Only datasets with persistent, resolvable and globally unique identifiers
(GUIDs) included, also known as persistent identifiers (PIDs), typically URIs
or handle/DOIs, also meaning that there is an existing landing page
dedicated for the associated metadata/data
8. FAIR maturity evaluations – approach
• Approach: randomly selection of ten data sets from each data
repository that passed the minimum requirements for digestion into
the FAIR Evaluator tool
• Machine-readable generic evaluations for F, A, I and R
• Recommendations are provided based on these evaluations, which
assists in identifying the parts where improvements can be made to
achieve a higher degree of FAIR
• Assistance in increasing FAIR maturities among a few selected data
repositories through a series of hackathons
• A repository FAIR improvement model will be the end result of the
FAIRification events
9. The main aims of doing FAIR evaluations
• To demonstrate what machine-actionability requires from data
repositories
• To set a baseline for FAIRness of data repositories
• To enable monitoring of FAIRness so that development efforts are
detected
• To monitor the evolution of FAIR metrics for data repositories in order
to trace the (positive) development of FAIRer research data and, if
possible, to gauge its effect on data reuse and scientific quality
10. The FAIR metric scores
• One score per letter (F, A, I, R), the total score and standard deviation
• No broad domain standards are implemented in the current version
of the evaluation
• Communities that have had their data repository evaluated should
take this as constructive feedback, and may use the results as
guidance in increasing the FAIRness of their service
11.
12. The 22 tests run by the FAIR Evaluator
Total FAIR score: 36.36 %. F= 37.50 % A= 40 % I= 42.85 % R= 0.00 %
13. Supporting data repositories on
standards
• Include requirements that are largely covered by the FAIR metrics,
including issues such as PIDs, licenses, data formats and metadata
standards
• Machine-actionable metadata are core to the FAIR Principles and will
require serious advancement toward the routine deployment of machine-
actionable metadata
• Approach: Demonstrating the methodology through Metadata for
Machines (M4M developed by GO FAIR) events
• Our support of “communities” is very likely going to be communities that
have certain commonalities, not scientifically, but rather a common FAIR
implementation profile
14. Supporting data repositories on
certification
• With the involvement of experienced data archives from the Nordics,
we expect to achieve a FAIR certification schema
• The schema will serve as a guide when assisting a handful of data
repositories in ultimately achieving a Core Trust Seal certification (or
DSA or lighter version) with additional criteria
• Due to limited resources and amount of time, we might focus on
assisting communities to perform self-assessments for CTS, or only
supporting them through parts of the certification process
• One physical workshop will be held where the participating
repositories can meet to raise questions and exchange experiences
15. About the FAIR Evaluator tool
• The maturity evaluations are in principle reproducible, although there is no version
control of how they are presented via the repository interface. Therefore, we archive the
evaluator output
• The harvester and the 22 indicators tests generate JSON output files that we store for
future reference and comparison
• In order to compare the results over time, we will run the same version of the harvester
and indicators at all times during the project
• The FAIR Maturity Evaluator tool takes as input a global unique identifier (GUID) and
explores the machine-actionable metadata provided within the dataset landing page
(whatever the GUID directs to). The existence of a GUID is the basic requirement for a
successful FAIR machine-readable evaluation
• Based on the successfully passed tests we provide ‘FAIR scores’ for each of the FAIR
letters based on the corresponding indicator tests that represent a given FAIR principle.
The FAIR Evaluator tool, developed by GO FAIR:
https://fairsharing.github.io/FAIR-Evaluator-FrontEnd/#!/
16. FAIRification of Nordic and Baltic data
repositories online event, April 22 2020
• Workshop for stakeholders of community data repositories in the Nordic and
Baltic region
• The goal is to provide guidelines and specific recommendations for organisations
hosting data repositories in order to maximise the reusability of research data
hosted by such entities
• The recommendations are based on the FAIR Maturity evaluations done in the
project of more than 100 repositories in the region
• The importance and added value of FAIRifying data repositories to adhere to
the FAIR principles will be discussed, as well as how the evaluations were carried
out using FAIR metric indicators
• Designed for research data repository managers, repository developers,
community stakeholders (e.g. data stewards)
• More information: https://www.eosc-nordic.eu/events/fairification-of-nordic-
and-baltic-data-repositories/