Guest post by Scholarly Repository Librarian Katherine Perdue and Research Data Management Librarian Joe Edgerton
In 2016, a group of academics, publishing and business professionals, and funding agency representatives worked together to help stem the tide of concern on making research data more sharable and easier to access. In a paper published that year, the authors established a set of principles they defined as “FAIR”: Findable, Accessible, Interoperable, and Reusable. These guiding tenants were developed, the authors wrote, to create “good data management” and to serve as “simple guideposts to inform those who publish and/or preserve scholarly data.”
It’s been 10 years since the FAIR principles were introduced, and quite a bit has changed since then. Open Research along with other “open” initiatives (e.g., Open Science, Open Data, Open Access) have continued to be championed, spawning new research platforms, policies, and cultural shifts in essentially every domain.
Now, in 2026, we are observing results of FAIR-related initiatives by seeing gradual adoption and recognition of the principles, according to the 2025 State of Open Data report. There are still many challenges facing researchers and their data, but now we are seeing researchers move beyond acknowledgement and into implementation and action.
The setup to FAIR
To better understand where we are today, we should travel back to the 2010s. It was around this time reports and surveys were coming out on how various disciplines’ research results were not reproducible. Related, the act of sharing underlying data and materials for published articles was not commonplace in most research fields. All this attention generated within the research communities called into question the process around publishing research and sharing data.
Even if research materials are made available at some point, they aren’t always easy for other people to work with. Perhaps a data table is placed in a static PDF image, or perhaps the meaning of some spreadsheet variable is not defined anywhere. Files could also be placed on a researcher's personal website with no easy way to reference them. Many of these issues around sharing data and digital files circle around machine actionability — how easy is it for a computer to work with or interpret a given file. We see the FAIR authors acknowledging, “ultimate machine-actionability occurs when a machine can make a useful decision regarding data that it has not encountered before,” anticipating the way AI agents and machine learning have become a larger part of research today.
What UVA has done with FAIR
UVA has incredible faculty and staff who have helped contribute to, and promote, FAIR principles. The late Phillip Bourne, founding dean of the University of Virginia School of Data Science, and Tim Clark, Associate Professor in the Department of Public Health Sciences and Neurology, were two of the original authors who helped design the core principles. They have both continued to advance the development and use of FAIR in multiple fields. At the University of Virginia Library, we also embody these principles in our work with researchers; one way we do this is through LibraData.
Ten Years of LibraData at UVA
Ten years ago in 2016, just as the FAIR principles were being codified, UVA Library embarked on its own mission to make UVA research data FAIR by launching LibraData.
LibraData is an open access data repository that uses the open-source Dataverse software platform developed at Harvard. When researchers deposit a dataset in LibraData, it becomes accessible worldwide on the internet. This means that if another researcher wants to replicate and verify a study, they have access to the raw data to do it. They can also reuse any datasets in LibraData as a basis for their own work, allowing the same data to be used for many different purposes beyond what was imagined by the original collector.
LibraData is FAIR
The key ingredient to making these datasets FAIR is the Digital Object Identifier (DOI). Upon deposit, each dataset receives its own DOI, a globally unique and permanent ID that can be used as a link. The DOI link will always take a user to the record for the dataset. Even if one day in the future UVA adopts different software and the dataset moves, the DOI link will still redirect to the right place.
DOIs are one part of making datasets Findable and Accessible, but they aren’t the only things making the system FAIR. LibraData allows users to search all datasets. It outputs metadata in multiple standardized machine-readable formats. Search engines, including specialized systems like Google’s Dataset Search, can read this Interoperable metadata, interpret the content, and deliver it to researchers who can learn from and Reuse it appropriately. The metadata includes subject specific fields to make it easier to find. It also specifies the license in a way that is legible to both human users and machines, so that it is clear to everyone how the data is allowed to be used.
LibraData’s mission to be FAIR requires constant improvements to keep up with changes in society and technology. For example, LibraData recently enabled a feature that enhances interoperability, adding Croissant, a metadata schema aimed at making data more comprehensible to AI, to all records. Another upcoming change will make it easier for users to link their datasets to their ORCID ID, which, like a DOI, is globally unique and permanent and makes it easier to find work by a specific researcher. Staying abreast of such updates will ensure that the datasets in LibraData stay relevant and findable long past their initial deposit.
LibraData Growth
When LibraData launched in March of 2016, it held just a few datasets in the areas of medicine, engineering, environmental sciences, biology, computer science, and the study of the humanities. Since then, it has grown enormously. At last count, 749 datasets, distributed between 51 collections, contained 53,712 individual files with over a million downloads between them. Just this year, 21 new datasets and one new collection have already been added, and we hope there will be many more to come.
Standing together, ready to help
At the Library, the Research Data Management and the Libra teams are here to support researchers as they navigate open and FAIR frameworks in the next decade and beyond. To make something FAIR takes a team effort, and you don’t have to do it alone. Please reach out to us!
Contact:
- The Libra Team, for any questions about depositing data, scholarly works, or other materials at UVA: libra@virginia.edu
- The Research Data Management team, for any questions about making your data FAIR, and data management in general: dmconsult@virginia.edu