Understanding the FAIR data principles and their use

Methodologies of scientific research have remarkably evolved to integrate a data-driven perception, rendering the data an indispensable asset that can acquire its own dynamism and be incorporated into a network of knowledge.

It hence becomes evident that the impact of research data has the potential to be significantly magnified. As researchers find and reuse previously produced data, knowledge becomes less insulated, and networks of resources emerge thus ensuring continuity and buildup of academic research.

In this context, the FAIR principles, corresponding to Findable, Accessible, Interoperable, and Reusable, were presented as the result of the work of the Data FAIRport Workshop in 2014.

How to share data can in fact be the subject of many restrictions, such as privacy and intellectual property. The FAIR data principles are not destined to override these restrictions, whether they are legal or ethical, but rather to fluidify and dynamize the movement of data through a better localization and traceability of data via metadata in a first step before the actual access.

These guiding principles are modulable and should be contextualized within the eventual specifics of a said discipline, legal framework, and institute. They present a way to make research more prolific and impactful, which explains their substantial integration into how the European Union is shaping its data strategy.

FAIR principles can benefit machine learning and researchers who have an interest in accessing the data. To comprehend how these principles, benefit research we must delve into how the data can be Findable, Accessible, Interoperable, and Reusable.

Content of the FAIR data principles

For it to be findable, the data should be subjected to the least obstacles possible that could hinder its localization. In this sense, metadata needs to be abundant for the raw data to be efficiently searchable. An identifier is thus important to link all the data processed in a certain context.

Accessibility entails that unnecessary obstacles to access should be lifted. This should be balanced with certain legitimate restrictions such as intellectual property or contractual non-disclosure clauses. The application of FAIR data principles does not lead necessarily to rendering the data open but rather making it easier to access through enriching the metadata.

Interoperability concerns how the data is presented, if the metadata is findable and accessible but is presented in a format that is not readable by a machine, the capacity of its flow is substantially hindered. A standardized format is therefore always needed to ensure that the data finds other uses. In this sense, standards can vary from one institution to another but efforts towards standardization are being pushed to, for instance, in the European Union.

Reusable entails the need for the data to showcase a set of details that organize and specify the possible utilization of the data. In this sense, for the data to be reused in a way that is aligned with intellectual property and other restrictions, licenses should be clearly defined as well as the stages of creation and previous reuse of the data.

The FAIR data principles in Privanova’s portfolio

In all projects in which Privanova handles aspects related to data management, it puts in place a methodology that ensures the respect of the four principles by all consortium members as they deal with the data they collect, produce, and process during the lifespan of the EU-funded project. These projects benefit from a Data Management Plan that covers a FAIR data principles component, such as DRG4FOOD, AI4HEALTHSEC, CYBERSPACE, GLOCALFLEX…

Privanova ensures that research outcomes of the EU-funded projects it is involved in are aligned with the FAIR data principles as they constitute a set of best practices that allows for research build-up. In parallel, the FAIR data principles allow the project to benefit from the dynamism that is offered by the data pooling in the European research communities.