Services um FAIR zu ermöglichen
Die FAIR-Prinzipien gelten für Metadaten und Daten. Ihre Umsetzung erfordert eine Reihe von Datenservices und -komponenten, die in einem breiteren Ökosystem dargestellt werden müssen. Diese Services selbst müssen auffindbar, identifizierbar oder indiziert sein und geeigneten Standards und Protokollen folgen, die Interoperabilität und Maschine-zu-Maschine-Kommunikation ermöglichen. Darüber hinaus bedarf es einer Reihe von Fähigkeiten im Bereich Datenmanagement.
Anmerkung: Die unten dargestellten Inhalte wurden aus aktuellen, vorrangig in Englisch vorhandenen Quellen zusammengestellt. Die Redaktion verzichtet hier bewusst auf eine deutsche Übersetzung.
A Data Steward should be the first person to whom researchers can turn for help. Specifically, in managing data, developing metadata and using repositories. A Data Steward helps to create mature working practices for Research Data Management (RDM) and according to FAIR principles.
FAIRsFAIR training events for Data Stewards is a good start to understand the existing competence frameworks for Data Stewards and other relevant professions which can be integrated on different education levels.
Rich Metadata
Metadata should be detailed and comprehensive, including descriptive information about the context and characteristics of the data. Rich metadata help a user/machine to understand how the data has been created, the fitness for other researchers’ demands, consistency with other datasets, and provide much of the content that the search engine will be able to index and discover.
In the following links you can find the most common catalogues and guidelines for developing metadata:
- RDA metadata standard
- Index of metadata standards
- The Data Documentation Initiative (DDI)
- Guide to writing „readme“ style metadata
Persistent Identifiers
Persistent identifiers are essential to robust data management strategies and enable the creation of trusted digital connections. Persistent identifiers eliminate the ambiguity in the meaning of the published data by assigning a unique identifier to each object, contributor, or organization.
Persistent Identifiers, a core component of connected research – ARDC will help you to determine what, why and when to use persistent identifiers (PIDs).
The following are the most commonly used PIDs for research data and for people:
Interoperability
A Data Steward should ensure data interoperability with applications for analysis and processing as well as data integration with other data. In this section we collect information related to standards in making data interoperable.
- Ontology vs Controlled Vocabularies
- OpenAIRE: Research graph data model
- Linked Open Vocabularies (linkeddata.es)
- Re3data.org , FAIRsharing & FAIRassist: Explore what resources exist — and if they can be used, extended or added new
- (5-star Open Data): Why Machine-actionability is core to each of the FAIR principles
FAIR data tools
How FAIR is our data? This is the most question that Data Steward been asked by researchers and ICT management!
Some highlights to find out whether data is aligned with FAIR, FAIRness assessment tools and some FAIR use cases:
- FAIR Toolkit and FAIR use cases for Life Science Industry
- Enabling FAIR Data – FAQs – COPDESS
- Assessment: FAIR Data and Software
Data licenses
Licenses and copyrights help to clarify the “R” in the FAIR principles. As Data Steward should know the tools and guides to licensing data and be able to help researchers to identify the owner of data:
Tools that help choose the right license:
- License Selector (ufal.github.io)
- Choose a License (creativecommons.org)
- Choose an open source license | Choose a License
- CLARIN License Category Calculator | CLARIN ERIC
Trustworthy repositories
Making data FAIR requires repositories with sustainable governance and organizational frameworks, reliable infrastructure, and comprehensive policies whilst preserving them over time. More information about the TRUST principles can be found on the Research Data Alliance (RDA) website.
- Recommendations on certifying services required to enable FAIR within EOSC: An analysis of activities relevant to certification of the services required to enable FAIR research outputs within EOSC
- Repository reflections on the FAIRsFAIR Repository Support Programme Part 3: Advice for repositories considering certification | FAIRsFAIR
- Practical Guide to the International Alignment of Research Data Management: A reference document on criteria for the selection of trustworthy repositories.
An ICT Operator is personnel who combines professional software engineering expertise with an intimate understanding of research1. They manage research infrastructures such as repositories.
FAIRsFAIR Repository Support Series Webinars | FAIRsFAIR is a good start to become familiar with FAIR-enabling practices
The following sections provide information that helps ICT Operators to identify and implement protocols, standards and languages in making data FAIR.
Data preservation
For data to be preserved it must be deposited in a repository and saved in file formats that will have the greatest utility in the future.
Below are some useful tools that support the preservation of digital objects and ensure their long-term usability:
- PREMIS: Preservation Metadata Maintenance Activity
- Metadata Encoding and Transmission Standard (METS)
Communication protocols
Protocols should be free (no need to pay to “use internet”) and open (specification of the protocol is known to everyone), which means they are globally implementable to facilitate data retrieval as defined in A1.2 ‘Anyone with a computer and an internet connection can access at least the metadata.’
- Communication protocols specifications:
- HTTP – Hypertext Transfer Protocol Overview
- FTP – File Transfer Protocol
Authentication and Authorization Protocols (retrievable protocols)
Allow a machine to automatically validate user identities and determine what operations and functions each user is permitted to perform. They could also alert the user to such requirements.
- Authentication and Authorization protocols specifications:
- Authentication and Authorization protocols implementation examples:
- OPI-PMH | Zenodo
- REST API | Zenodo
PID registration bodies
Vocabularies and languages
For knowledge representations that provide a mechanism to make data machine-actionable:
- DCMI: DCMI Metadata Terms
- OLS: Documentation < Ontology Lookup Service>
- Rocrate
- An inventory of tools for converting your data to RDF
How systems communicate with each other:
- DCAT
- org – Schema.org
- Optimization of search engine – Sitemap: Search engine optimization
This group refers to the management of institutions/organizations that perform research activities. They define data policies and govern procedures relevant to FAIR principles within their institutions/organizations.
Research funders and publishers who encourage the generation of FAIR data could also find some useful references here.
The following sections provide information that help ICT management enabling FAIR principles across their institutions/organizations.
Implementing FAIR in research data policies within universities
- European Code of Conduct for Research Integrity Research Integrity
- Turning FAIR into reality (europa.eu)
- Policy Makers | FAIRsFAIR
- Guidelines on FAIR Data Management | Horizon 2020
- List of social science journals with a data availability policy
Good practices in FAIR education and training
- Dutch roadmap towards national implementation of FAIR data stewardship
- Recommendations on practice to support FAIR data principles | FAIRsFAIR
- FAIR Adoption Handbook | FAIRsFAIR
- FAIR Data Maturity Mode| RDA
Institutional challenges and drivers in implementing FAIR
Remember!
- „As open as possible, as closed as necessary“.
- Even sensitive and private data can be FAIR! (FAIR for Sensitive Data).
Tip
Find here the video recording and presentation „FAIRification of data and services“ (EOSC Future Provider Days, April 2022).
FAIR Cookbook
The FAIR Cookbook contains ten recipes for the Life Science domain that provide practical support for researchers, data stewards, trainers, and developers on how to FAIRify data, assess FAIRness, which models, technologies, tools and standards, as well as the required skills to achieve and improve FAIRness. Also interesting for other disciplines!
FAIR Use Cases
Zenodo offers an overview of how the service responds to the FAIR principles.