Introduction

  • FAIR means making research objects more usable by humans and machines, not automatically making them open.
  • Machine-friendly objects are structured so software can interpret and reuse them reliably.
  • Digital objects include datasets, publications, metadata records, and related documentation.
  • DOI is a common PID used for datasets and publications.

Set up your own terms

  • A data terms of use statement defines the legal and practical basis for reuse.
  • A license is the minimum requirement, but some projects need richer terms or a custom agreement.
  • Store terms of use in an accessible text format such as .md or .txt.
  • If a standard license does not fit the project, a tailored terms-of-use statement or usage agreement may be necessary.

Speak the same language

  • Data descriptions may appear under names such as codebook or data dictionary.
  • Reusing ontology terms reduces ambiguity and improves interoperability.
  • A useful description links dataset variables to accepted community concepts.
  • Linked data becomes more feasible when descriptions are structured and identifier-based.

Securely share

  • Data access protocols describe how humans and machines gain access to data.
  • Open access is an access model that can still be expressed through explicit rules and interfaces.
  • Human-facing request forms and machine-facing APIs are both important access patterns in FAIR data practice.
  • FAIR service endpoints are more useful when they are registered and discoverable.

Publish and preserve

  • Repositories improve preservation, discovery, and citation.
  • Prefer a trusted community repository when one fits your discipline.
  • General repositories such as Zenodo still provide a strong baseline for FAIR publication.
  • DOI is a key persistent identifier for datasets and publications.

Make machines work for you

  • Rich metadata combines descriptive metadata with shared vocabularies and a structured machine-readable format.
  • JSON-LD is a common way to publish rich metadata.
  • Repositories can often generate rich metadata automatically.
  • Rich metadata should be published anywhere the digital object is stored or represented.

Responsibly reuse

  • Reuse starts with clear citation and persistent identifiers.
  • Rich metadata is essential for search engines and aggregators to discover datasets.
  • Human-friendly web pages are not enough; machine-readable structure matters.
  • Testing discoverability is a practical part of making data reusable.