Understanding the FAIR Principles

The continuous growth of the Open Science movement raises some challenges. One of them is to enable knowledge discovery and a more satisfactory reuse of research data and methods. In this context, various stakeholders (researchers, scholarly publishers, funding agencies and industry representatives) jointly designed a concise and community-agreed set of foundational principles: Findability, Accessibility, Interoperability, and Reusability (Martone, 2015). Wilkinson et al. (2016) formally published the FAIR principles for the first time as “The FAIR Guiding Principles for scientific data management and stewardship”, in the journal Scientific Data.
“FAIR refers to a set of principles, focused on ensuring that research objects are reusable, and actually will be reused, and so become as valuable as is possible” (Mons et al. 2017). The principles–Findability, Accessibility, Interoperability, and Reusability–do not specify technical requirements; instead, they are a set of high-level, concise, domain-independent, guiding principles. Mons et al. (2017) states that “FAIR simply describes the qualities or behaviours required of data resources to achieve – possibly incrementally – their optimal discovery and scholarly reuse.”
The different elements that constitute the principles are related but independent and separable. By means of diverse implementations, they make possible “a continuum of increasing reusability” (Mons et al. 2017). As a guide, FAIR helps different research actors (producers, publishers, stewards) to comply with the requirements of open science and open data. They also assist the actors, especially publishers and stewards, in assessing their implementation choices and the degree of “FAIRNess” of their data.
FAIR’s ultimate goal is the reuse of research objects by improving the ability of machines to automatically find and use the data, besides supporting its reuse by individuals. The table below details the FAIR principles: Findable (F1, F2, F3, F4), Accessible (A1, A1.1, A1.2, A2), Interoperable (I1, I2, I3), and Reusable (R1, R1.1, R1.2, R1.3).
The FAIR principles in detail
| FAIR Principle | F1 |
| Definition | (Meta)data are assigned a globally unique and persistent identifier |
| Commentaries | Identifiers are fundamental for human-machine interoperation.Globally unique means that every time an identifier is reused/reassigned, it has to be done with reference to the data. Persistent means the link must remain available at least to some degree even if the resource is no longer available.Identifiers may be produced by internal or external registry services.Examples: DOI, ORCID |
| FAIR Principle | F2 |
| Definition | Data are described with rich metadata (defined by R1 below) |
| Commentaries | One should be able to find data based on the information provided by their metadata, even without the data’s identifier. Metadata should be generous and extensive.It should include descriptive information about the context, quality and condition, or characteristics of the data. |
| FAIR Principle | F3 |
| Definition | Metadata clearly and explicitly include the identifier of the data they describe |
| Commentaries | The metadata and the dataset they describe are usually separate files. The association between a metadata file and the dataset should be made explicit by mentioning a dataset’s globally unique and persistent identifier in the metadata. Example: The connection should be annotated in a formal manner in case of RDF metadata |
| FAIR Principle | F4 |
| Definition | (Meta)data are registered or indexed in a searchable resource |
| Commentaries | (Meta)Data should be findable. The availability of a digital resource such as a dataset, service or repository should be known so that the resource can be discovered. Indexing is a way to make digital resources discoverable. |
| FAIR Principle | A1 |
| Definition | (Meta)data are retrievable by their identifier using a standardized communications protocol |
| Commentaries | FAIR data retrieval should be mediated without specialized or proprietary tools or communication methods. This principle focuses on how data and metadata can be retrieved from their identifiers.Barriers to access such as protocols that have limited implementations, poor documentation, and components involving manual human intervention should be avoided.In case of highly sensitive data, it is possible and FAIR to provide a form of contact to discuss access to the data. This contact protocol must be clear and explicit in the metadata.Example: http(s) or ftp. |
| FAIR Principle | A1.1 |
| Definition | The protocol is open, free, and universally implementable |
| Commentaries | The protocol should be free and open-sourced and thus globally implementable to facilitate data retrieval. Anyone with a computer and an internet connection can access at least the metadata. Examples: HTTP, FTP, SMTP, telephone, do not use Skype and Microsoft Exchange Server protocol because they are proprietary |
| FAIR Principle | A1.2 |
| Definition | The protocol allows for an authentication and authorisation procedure, where necessary |
| Commentaries | In case of heavily protected and private data, accessibility requires authentication.Example: HMAC authentication, HTTPS, FTPS, Telephone |
| FAIR Principle | A2 |
| Definition | Metadata are accessible, even when the data are no longer available |
| Commentaries | Metadata are valuable in and of themselves, when planning research, especially replication studies. Even if the original data are missing, tracking down people, institutions or publications associated with the original research can be extremely useful.Metadata must remain available even when the data is no longer sustained. |
| FAIR Principle | I1 |
| Definition | (Meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation. |
| Commentaries | Data should be readable for machines without the need for specialised or ad hoc algorithms, translators, or mappings. To ensure automatic findability and interoperability of datasets, it is critical to use: (a) commonly used controlled vocabularies, ontologies, thesauri (having resolvable globally unique and persistent identifiers, F1); and (b) a good data model (a well-defined framework to describe and structure (meta)data).Example: The RDF extensible knowledge representation model is a way to describe and structure datasets, OWL, DAML+OIL, JSON LD |
| FAIR Principle | I2 |
| Definition | (Meta)data use vocabularies that follow FAIR principles |
| Commentaries | The controlled vocabulary used to describe datasets needs to be documented and resolvable using globally unique and persistent identifiers. This documentation needs to be easily findable and accessible by anyone who uses the dataset. |
| FAIR Principle | I3 |
| Definition | (Meta)data include qualified references to other (meta)data |
| Commentaries | A qualified reference is a cross-reference that explains its intent. Specify if one dataset builds on another data set, if additional datasets are needed to complete the data, or if complementary information is stored in a different dataset. In particular, the scientific links between the datasets need to be described. |
| FAIR Principle | R1 |
| Definition | (Meta)data are richly described with a plurality of accurate and relevant attributes |
| Commentaries | A user (machine or human) should be able to decide if the data is actually useful in a particular context. Labels should be attached to the data.The metadata provided should allow discovery. Also metadata should richly describe the context under which the data was generated (ex. experimental protocols, the manufacturer and brand of the machine or sensor that created the data, the species used, the drug regime, etc.). |
| FAIR Principle | R1.1 |
| Definition | (Meta)data are released with a clear and accessible data usage license |
| Commentaries | ‘I’ covers elements of technical interoperability. R1.1 is about legal interoperability.Usage rights attached to the data, licensing status, and the conditions under which the data can be used should be clear to both machines and humans.Examples: MIT or Creative Commons |
| FAIR Principle | R1.2 |
| Definition | (Meta)data are associated with detailed provenance |
| Commentaries | Data provenance should be clear (i.e., clear story of origin/history), as well as who to cite and/or how the author wishes to be acknowledged. A description of the workflow that led to the data should also be included. |
| FAIR Principle | R1.3 |
| Definition | (Meta)data meet domain-relevant community standards |
| Commentaries | Community standards or best practices for data archiving and sharing should be followed if they exist.It is easier to reuse data sets if they are similar: same type of data, data organised in a standardised way, well-established and sustainable file formats, documentation (metadata) following a common template and using common vocabulary. |
Source: GO FAIR
References
Featured Image: FAIR data principles by under CC BY-SA 4.0 license
Martone, Maryann E. 2015. “FORCE11: Building the Future for Research Communications and e-Scholarship”. Bioscience 65 (635). https://doi.org/10.1093/biosci/biv095
Mons, Barend, Cameron Neylon, Jan Velterop, Michel Dumontier, Luiz Olavo Bonino da Silva Santos, and Mark D. Wilkinson. 2017. “Cloudy, Increasingly FAIR; Revisiting the FAIR Data Guiding Principles for the European Open Science Cloud.” Information Services & Use 37 (1): 49–56. https://doi.org/10.3233/ISU-170824.
Wilkinson, Mark D., Michel Dumontier, IJsbrand Jan Aalbersberg, Gabrielle Appleton, Myles Axton, Arie Baak, Niklas Blomberg, et al. 2016. “The FAIR Guiding Principles for Scientific Data Management and Stewardship.” Scientific Data 3 (1): 160018. https://doi.org/10.1038/sdata.2016.18.
OpenEdition suggests that you cite this post as follows:
Karla Avanço (June 8, 2021). Understanding the FAIR Principles. The road to FAIR. Retrieved June 12, 2026 from https://doi.org/10.58079/trro
1 Response
[…] animate the discussion, CO-OPERAS IN would like to know how it can help you apply the FAIR principles in your research processes. Use the commentary box below to share your doubts, remarks, suggestions […]