FAIR principles and Open Science
The FAIR festival organized by GOFAIR opened on June 21st with a plenary session named: “FAIR minds open science”. Represented by Elena Giglia, CO-OPERAS had the chance to participate directly to the discussion, but more importantly, the session allowed to investigate the strong connection between the Open Science movement and the FAIR principles. You can listen to the panelists’ own visions on the topic in the video available here: https://www.go-fair.org/resources/go-fair-materials/media/.
The session seemed a good opportunity to review some of the main notions related to Open Science and… the FAIR principles. As it is often said, “FAIR doesn’t mean open”. However, as reminded by Jean-Claude Burgelman during the plenary, the FAIR principles are essential to Open Science, and more specifically, to the building of the European Open Science Cloud.
Simply put, the FAIR Principles represent the common technical layer that enables the construction of an appropriate digital ecosystem for research. The FAIR Principles also allow for modulating the openness of FAIRified objects, as long as they are “as open as possible, as closed as necessary”.
An innovative set of rules distinct from previous initiatives, the FAIR principles complement the Open Science movement. We list below a few notions related to Open Science, specifying their similarities and differences with the FAIR principles.
Other notions related to Open Science will be presented in future posts.
The open definition by the Open Knowledge Foundation was released in 2005. It sets out precise principles to define openness in relation to data and content. It provides a standard that is important because the value of open data relies on the quality and the guarantee of its legal and technical compatibility with other sources. In this context, two notions appear to be crucial: open science and open data.
Open science is a growing movement to make scientific processes more transparent and publications and data more available. It “aims to build an ecosystem in which science will be more cumulative, more supported by data, more transparent, faster and providing universal access” (Ouvrir la science, translation by the author). This movement has been supported by technological developments and legal advancements, which leads to the availability of a considerable amount of scientific and technical information on the web.
Open science means to practice science according to open principles: open access publishing, open data, and open source (Figure 1).
Open data refers to data that can be freely accessed, used, modified, and shared. It should be legally and technically open. Legally open means the data are available under an open license that allows anyone to freely access, reuse, and redistribute these data. Different types of open licenses must be compatible with each other and satisfy some conditions such as required permissions, modification, separation, no charge, application to any purpose, among others. Technically open means that the data are available in machine-readable form and for no more than the cost of reproduction.
Machine-readable is a crucial notion for both open data and FAIR. It refers to the data format that can be automatically read by a computer, such as CSV, JSON, and XML. The data must also be structured. In other words, the structural relation between the elements that constitute the data must be explicit.
A few principles, summarized below (Figure2), constitute the foundation of open data.
The notion of open data should be well understood, as the Open Data Handbook emphasizes, because of interoperability, that is, the ability to interoperate different datasets, allowing various components to work together. It is noteworthy that interoperability is one of the requirements of FAIR; however, FAIR data does not necessarily mean open or free. Mons et al. (2017) states that the “FAIR principles, although inspired by Open Science, explicitly and deliberately do not address moral and ethical issues pertaining to the openness of data.” The author continues by saying the following:
“FAIR only speaks to the need to describe a process – mechanised or manual – for accessing discovered data; a requirement to openly and richly describe the context within which those data were generated, to enable evaluation of its utility; to explicitly define the conditions under which they may be reused; and to provide clear instructions on how they should be cited when reused. None of these principles necessitate data being ‘open’ or ‘free’” (Mons et al., 2017).
The FAIR principles require to comply with accessibility and especially reusability, as well as transparency about access and reuse conditions. Furthermore, they require a clear, preferably machine-readable, license. Finally, the “transparent but controlled accessibility of data and services, as opposed to the ambiguous blanket-concept of ‘open’, allows the participation of a broad range of sectors – public and private – as well as genuine equal partnership with stakeholders in all societies around the world.” (Mons, 2017).
FAIR speaks to the need of describing, richly and openly, the context of data production to allow the assessment of its utility. Furthermore, it is about defining the conditions and providing instructions for its reuse.
To advance the discussion
The Project Horizon 2020 states that participating in the Project does not necessarily mean opening up all the research data. Rather, it follows the principles “as open as possible, as closed as necessary” and encourages sound data management as best practice in research.How this principle applies to different stakeholders coming from different contexts is subject to reflection.
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.
Featured image: Pixabay license.