Source: own illustration
The FAIR principles are a fundamental concept in research data management (RDM). They were first published ten years ago in the journal *Scientific Data* and have since established themselves as an international standard for the sustainable management of research data. The RDM services at RWTH Aachen University are also based on the FAIR principles and thus support researchers in ensuring their research data is FAIR throughout the entire research process.
We are taking this anniversary as an opportunity to take a closer look at the FAIR principles and their origins in this blog post.
Origins of the FAIR Principles
The FAIR principles were first described in 2016 in the article The FAIR Guiding Principles for Scientific Data Management and Stewardship by Marc D. Wilkinson et al. In it, the authors address the question of how to ensure the most effective possible reuse of digitally available research data. The article places particular emphasis on the readability of data by machines with little or no human intervention. Wilkinson et al. summarize their findings with the acronym FAIR: FAIR stands for Findable, Accessible, Interoperable, and Reusable.
F – Findability
The aspect of findability deals primarily with metadata. Clear metadata helps both humans and machines to quickly find and categorize data. To further increase findability, metadata should be entered into searchable databases.
In addition to detailed metadata, so-called Persistent Identifiers (PIDs) also contribute to the findability of research data. PIDs are permanent and unique identifiers for objects, such as datasets or journal articles. The best-known PID is the Digital Object Identifier (DOI), which has become a standard in the scientific community. Individuals can also be identified with a PID: the Open Researcher and Contributor ID (ORCID) is used for this purpose.
At RWTH, various FDM services support the discoverability of research data. Coscine supports the implementation of this FAIR principle by automatically assigning PIDs at the project and resource levels. Researchers can also log in to Coscine using their ORCID and thus uniquely associate their research data with their researcher identity. RWTH Publications is also available for publishing datasets or other research results, ensuring that data can be discovered and cited over the long term.
A – Accessibility
Another component of the FAIR principles is data accessibility. This aspect is often misunderstood. Accessibility does not necessarily mean that all data must be available free of charge and without restrictions. While such absolute public availability is an ideal in the spirit of open science and should be pursued wherever feasible, in reality there are legal restrictions that prevent this.
Rather, the aspect of accessibility means that the metadata clearly communicates the conditions under which the data is accessible and the steps that must be taken to view and use the data. This means that even personal data or data otherwise protected by privacy laws can be FAIR.
In Coscine, researchers can individually control access to their research data using the Owner, Member, and Guest roles. Additionally, they can specify whether a project’s metadata is visible to the public or exclusively to project members. This allows even research data that is not publicly accessible to be managed in accordance with FAIR principles.
I – Interoperability
To ensure data compatibility, the relationships between datasets should be made clear. If one dataset refers to or builds upon another, this should already be evident from the metadata.
The language in which data is stored is also important for data compatibility. It should be readable by both humans and machines and follow a clear logic.
Coscine also supports the interoperability of research data through standardized, machine-readable metadata profiles based on established standards. In addition, controlled vocabularies and subject-specific terminologies can be used to describe research data in a consistent manner. Relationships between datasets can also be documented, allowing them to be linked to one another and processed by different systems. The metadata profiles used are managed, among other places, via the AIMS platform, which promotes their harmonization and facilitates the exchange of research data across disciplinary and institutional boundaries.
R – Reusability
Metadata is once again relevant to reusability. While the aspect of discoverability is primarily about making the data as easy to find and clearly identifiable as possible, reusability focuses on describing the contents of the dataset in the metadata. In addition to the content, the metadata should also indicate, among other things, whether the data is raw or processed and what method was used to collect it. This allows researchers to better assess whether the data is useful to them without having to examine the entire dataset directly. Furthermore, the metadata should clearly specify the licenses and usage rights under which the data may be used.
At RWTH, various FDM services support the reusability of research data:
As early as the planning phase, researchers can use the Research Data Management Organizer (RDMO) to centrally record important information on data management in so-called data management plans and develop strategies for later reuse.
During the research process, the Electronic Lab Notebook (ELN) eLabFTW helps researchers document experiments and workflows in a structured manner.
With Coscine, research data can be described in terms of content using comprehensive metadata profiles and assigned usage licenses, ensuring that important information regarding the data, its origin, and conditions for reuse is available.
For the publication of research data and other scientific results, RWTH Publications offers a way to make them accessible and citable in the long term.
FAIR Principles – 10 Years Later
Ten years after the publication of the groundbreaking article by Wilkinson et al., the FAIR principles have become indispensable in research data management. For example, the European Union promotes and requires FAIR practices in research data management (FDM) as part of the FAIRSFAIR project. The German Research Foundation (DFG) also incorporates them into its guidelines for ensuring good scientific practice and has included the FAIR principles in its Code for Ensuring Good Scientific Practice.
At RWTH Aachen University, as described above, the FAIR principles form the foundation of its research data management (RDM) services: RDMO enables researchers to plan their data management; eLabFTW supports the structured documentation of experiments and workflows; Coscine facilitates the organization and storage of research data with comprehensive metadata; and RWTH Publications supports the publication of research results. This offering is complemented by the IT Center’s FDM consulting services, which support researchers in implementing the FAIR principles throughout the entire data lifecycle.
The FDM team will be happy to assist you in implementing the FAIR principles for your research data and answer any other questions you may have about FDM.
Responsible for the content of this article are Hedda Faber, Katharina Grünwald and Arlinda Ujkani.
The following sources served as the basis for this article:
Coscine: Umsetzung der FAIR-Prinzipien mit Coscine
DFG: Leitlinien zur Sicherung guter wissenschaftlicher Praxis
Leave a Reply
You must be logged in to post a comment.