Metroline Step: Define FAIRification objectives

Owned by Sander de Ridder
Last updated: Sept 27, 2024 by Hannah Neikes
5 min read

statuS: Ready for review

Short description

‘A FAIRification objective is the end goal that the owners of a resource [e.g. dataset, database, protocols, analysis workflow] are looking to achieve with the process of FAIRification. It needs to be based on the research question, the data available in the resource, the expertise of the FAIRification team, available FAIR solutions and the availability of budget and appropriate infrastructure.’ (EJP RD FAIRification Guidelines)

Defining FAIRification objectives is the starting point of your FAIRification journey. Spend some time considering your FAIR objectives and the reasons behind them. What is your FAIR driving force? It is a fundamental step in making data FAIR, according to FAIRifcation models, such as FAIRopoly, A Generic Workflow for the Data FAIRification Process, FAIR in Action Framework by FAIRplus, A goal-oriented method for FAIRification planning and EJPRD FAIRification objectives. These FAIR objectives may be set by funders, journals or (research) organisations (see FAIRsharing), who encourage data FAIRness to stimulate reuse of data.

FAIRification objectives can focus on (see FAIR Cookbook):

  • a single aspect of FAIR, such as increasing Findability of a dataset by making it available in the Dutch National Health Data Catalogue;

  • multiple aspects of FAIR, such as use of terminologies and vocabularies to represent the data (Interoperability) combined with making the dataset available in the before mentioned National Catalogue (Findability).

Why is this step important

Based on the aforementioned models, setting clear FAIRification goals is essential for a number of reasons, some of which are described below.   

  • Clarity of purpose. It provides a clear purpose and direction for FAIRificaton efforts in (clinical) studies and projects. It helps understanding who is responsible for what and why it matters.

  • Collaboration and communication. Team members and others can communicate more easily when FAIR objectives are clearly defined. It guarantees that all parties involved are aware of the FAIR goals and are able to collaborate to meet them.

  • Measurable progress. Clearly defined FAIR objectives act as measurable milestones and guarantee that FAIRification efforts have the biggest possible effect. They support monitoring the procedure and identifying areas for improvement.

Importantly, clear FAIR objectives help identifying which steps of the FAIR Metroline are relevant to your FAIRification journey.

How to

Step 1

Get acquainted with the FAIR principles, for instance via the GO FAIR Foundation (GFF) website. These form the basis for defining FAIR objectives.

Step 2

When defining your FAIR objectives, check if any of the following applies.

  • What are the FAIR requirements set by the funder of your project or the journal you aim to publish in? FAIRsharing provides an overview of policies.

  • What are the FAIR requirements set by your institute? For instance, uploading your (meta)data to a certain repository or catalogue.

  • What are the FAIR requirements set by your project? These are usually written down in your project’s Data Management Plan (DMP). DMP tooling, such as Data Stewardship Wizard (DSW), help to make your study or project more FAIR.

Step 3

For onboarding metadata in the Dutch National Health Data Catalogue, the following FAIR objectives are relevant.

  • The (meta)data should be Findable by humans and computers in the Catalogue (Findability).

  • Information should be provided on how the data can be accessed. Note: accessible does not imply Open. Data should be “As open as possible, as closed as necessary” (Open Science in Horizon 2020).(Accessibility).

  • Provide your metadata via a FAIR Data Point to the Catalogue. For this the metadata must be mapped to the DCAT-AP based Catalogue’s core metadata schema (Interoperability).

  • The data should be provided with a machine-readable license and provenance information on how the data was formed (Reusability).

Step 4

You may have your own requirements to formulate FAIR objectives, for example:

  • I would like to add more metadata to improve the reusability of my dataset;

  • I’m setting up a new data collection and my data must be collected in a standardised, interoperable way;

  • I want to transform my existing data into an interoperable dataset;

  • I have standard, core data which needs to be collected and made FAIR in real time.

Step 5

Decide on the desired FAIR level for your project or study, using resources such as the RDA FAIR Data Maturity Model, FAIRplus Dataset Maturity (DSM) Model or GFF's FIP mini-questionnaire. Does a lightweight FAIR layer suffice, i.e. improving from non-structured to structured metadata? Or should it be more extensive, i.e. transforming a relational database to fully ontologised linked data?

See Metroline Step Pre-FAIR assessment for more information.

Step 6

Identify the resources (team of experts, time, tools, etc.) you need to complete your FAIRification process and whether they are available. For instance:

  • is there someone in your team that has run a FAIRification project before?

  • do you need to hire specific expertise?

  • do you need specific software or additional hardware, that allow you to meet your FAIR objectives?

See Metroline Step: Build the Team for more information on team requirements.

Step 7

Consider the impact of your FAIRification efforts. For instance:

  • Cost-benefit ratio. Does the end result outweigh the efforts you are putting in it?

  • Reuse of data. In what way do potential reusers of your data (e.g. researchers, other users of the data) benefit most from the results of your FAIRification efforts?

  • Stakeholder benefits. What other stakeholders (e.g. patients, general population, clinicians) would benefit from making your data more FAIR?

  • Research question. How does the driving research question impact your FAIRification objectives?

Step 8

Organise and describe the steps and resources needed for reaching your FAIRification objective(s). This can be a separate plan (see Metroline Step: Design Solution Plan - in development) or part of your DMP. Make sure to formulate your FAIR objectives as specific as possible and, where possible, incorporate research questions. For example:

  • I require Interoperable data, because I need to connect two different datasets to answer question X;

  • I want to make my data more Findable for researchers in my field by publishing it in repository Z.

Expertise requirements for this step 

Defining FAIRification objectives is typically a collaborative effort by a range of experts, as described in Metroline Step: Build the Team.

  • Researchers with domain knowledge. Provide context to the FAIRification efforts from the perspective of a domain.

  • FAIR data stewards. Help defining FAIR objectives to meet the project’s, funder’s, journal’s and/or institute’s requirements.

  • Semantic experts. Help specifying the metadata/modelling aspects of FAIR objectives.

  • ELSI experts. Help identifying the legal compliance and ethical aspects of FAIR objectives.

Practical examples from the community 

European Joint Programme on Rare Diseases (EJP RD)
Programme aiming to create an effective rare diseases research ecosystem.

FAIR objectives include, for example (see for more information the FAIRopoly website):

  • an interoperable registry, requiring standardised and machine-readable data;

  • identification of patient cohorts for clinical trials, requiring to collect standardised and reusable data;

  • a Rare Diseases research ecosystem, requiring data to be completely FAIR so that information can be queried at the source.

Metachromatic leukodystrophy initiative (MLDi)
The MLDi is an international patient registry for MLD and an academic collaborative network

FAIR objectives include (see also the MLDi demonstrator project page):

  • sustainable MLDi registry, requiring interoperability with future and existing (international) MLD databases;

  • reusable MLDi data, by creating a web-based semantic model to improve interoperability;

  • FAIR as team effort, requiring people with the right expertise.

Training

Relevant training will be added in the future if available.

Suggestions

Visit our How to contribute page for information on how to get in touch if you have any suggestions about this page.