FAIR Implementation Profiles (FIPs) and FAIR Enabling Resources (FERs) as accelerators of Interoperability and Reusability of Datasets across Multiple Communities.

The Findable, Accessible, Interoperable, Reusable (FAIR) Guiding Principles of Wilkinson et al. (2016) put specific emphasis on using 'rich' metadata to enhance the ability of both humans and machines to find, access, aggregate and reuse data. Yet many datasets labelled as FAIR-compliant only meet Findable and Accessible and then only for humans. Few datasets are Interoperable and Reusable by machines mainly because there is insufficient specificity and granularity on the exact standards, formats and vocabularies implemented. FAIR Implementation Profiles (FIPs) enable a community to declare the implementation choices it makes for each FAIR principle. Each choice is then published as a FAIR-Enabling Resource (FER), which is a machine-actionable nano-publication of any object (vocabularies, services, metadata profiles, etc) used to comply with a FAIR principle. FIPs, combined with FERs, can accelerate machine-actionable interoperability and reusability of datasets acquired by different communities, and at scale. FIPs can be visualised as knowledge graphs to help identify commonalities across communities, and provide more specificity than existing FAIR assessment methodologies. The CODATA and RDA-led WorldFAIR Project on Global Cooperation on FAIR Data Policy and Practice (https://worldfair-project.eu/), funded by the European Commission WIDERA Programme, reviewed different approaches to FIPs from eleven disciplinary case studies: Chemistry, Nanomaterials, Geochemistry, Social Surveys, Population Health, Urban Health, Biodiversity, Agricultural Biodiversity, Oceans, Disaster Risk Reduction and Cultural Heritage. This paper will highlight how the Geochemistry case study created FIPS at multiple community levels (repository, data collection, dataset and reference) and will then compare and contrast with approaches of the other disciplines.