Introducing the FAIR Principles for research software

Introducing the FAIR Principles for research software
Like

About the Article

The newly published article, Introducing the FAIR Principles for research software, provides information about a new set of principles that can improve the findability, accessibility, interoperability, and reusability (FAIR) of research software. This work is an extension of the original FAIR Guiding Principles. While these original principles were aimed at all research outputs, in practice they were developed specifically for research data and in the context of the research data ecosystem. Because research software is not just another kind of research data, but has distinct properties and a distinct ecosystem in which it is created, used, and distributed, distinct principles are needed to make it FAIR. This work was done under three organisations: Research Software Alliance (ReSA), Research Data Alliance (RDA), and FORCE11, in order to garner substantial community input during the development process and to ensure wide dissemination and uptake of the final principles.

Key Findings

The process of developing FAIR principles for research software was based on understanding the differences between research software and research data, and how they are created, used, and distributed. These difference led to an overlapping but distinct set of principles, where the differences are based on the characteristics of open source software (e.g., hierarchical from project to version to files to lines of code, developed in the open, able to be forked and merged, shared via social coding platforms and package management systems) as well as differences in the meaning of interoperability (how software interacts with other software) and reusability (the ability to read the code, to build it, to modify it, and to execute it.)

Additionally, a shortcoming of the FAIR data principles has been the lack of a governing organisation, which has been somewhat addressed by the emergence of GO-FAIR. The FAIR principles for research software have explicitly been versioned (currently 1.0) and the RDA, through its Software Source Code Interest Group, has been defined as the group to interpret the FAIR4RS principles and to revisit them in 2 years to determine if any updates are needed.

About the Methodology

The work to create a set of principles for FAIR research software began with the nine leaders of the FAIR4RS Working Group coming together, and publicising the idea via a session at the RDA plenary, and via open community calls advertised through the three organising agencies. The working group then divided the work to be done into understanding past work, taking a fresh look at changes needed in the FAIR Guiding Principles, and defining research software, and then came together again with a partial draft set of principles and a set of open questions. Community response to the draft and the questions led to a second draft set of principles, and after more community consultation and additional subgroups looking at adoption, adaptation, and future governance, the final set of principles were approved and published. Overall, the FAIR4RS Working Group engaged about 500 people (from 110 organisations over 34 countries) in the development of the principles, including the more than 240 FAIR4RS WG members. The methodology has been described in a separate paper, The FAIR4RS team: Working together to make research software FAIR.

About the Authors

Dr. Barker, Prof. Chue Hong, and Dr. Katz are primary co-authors who equally led the writing, reviewing, and editing, with the remaining authors contributing equally to the writing and reviewing. All authors were part of the leadership group managing development of the principles.



Please sign in or register for FREE

If you are a registered user on Research Data at Springer Nature, please sign in

Go to the profile of Mark Musen
8 days ago

It is time for research software to be treated as a first-class digital research object.  The FAIR principles described in this paper provide a means for technologies that have been applied to research data to help to make research software FAIR in a completely analogous manner.