An Hourglass approach to an Internet of FAIR Data and Services
FAIR is about automating the Findability, Accessibility, Interoperability and Reusability of digital resources populating the Internet. Although the formulation of FAIR Principles occurred relatively recently (2016), numerous infrastructure frameworks have been proposed over the last two decades that anticipated and, in many cases demonstrated, behaviors described by the FAIR Principles.
Among these frameworks are the Digital Object Architecture (first proposed by Kahn and Wilenski and more recently developed by the GEDE Working Group of RDA) which proposes abstraction layers for automated data operations and IBM’s Linked Data Platform, later to become a W3C standard, harnessing Semantic Web approaches emphasizing operations that are contingent on semantically rich (domain-relevant) descriptions.
As these and other approaches have demonstrated key capabilities, it is increasingly recognized that the barriers to the emergence of a global data infrastructure is no longer technological, but social: finding widespread consensus on particular implementations. It is well-known that large integrative infrastructures demonstrate widespread uptake through an “hourglass” strategy, where “convergence” is driven by absolute minimal standards on the most generic operations, leaving maximum freedom to implement solutions in specialized domains and in use cases involving legacy data.
Beginning in August 2019 GO FAIR and others coordinated a series of international “convergence” meetings that brought together key stakeholders in Digital Object Architecture and Semantic Web, to assess the FAIR profiles of each, and where possible, to look for complementary features. Guided by the FAIR Principles, it was possible to sketch a novel approach that combines the powerful features from both Digital Objects and Linked Data yielding the FAIR Digital Object Framework. The FAIR Digital Object has the potential to serve as the “center of the hourglass” for an emerging Internet of FAIR Data and Services. It is giving users a definite but minimal standard to build toward, yet also the freedom to choose their own implementation paths. By adhering to the FAIR Principles in all design elements, the FAIR Digital Object guarantees FAIRness for even complex multicomponent digital artifacts.
An international consortium is now actively working together to formulate a global consensus on a specification for the FAIR Digital Object, and to demonstrate multiple implementations. If you would like to join this effort, please consider adding your name to the Joint Statement on FAIR Digital Object Framework.
- August 28-29, 2019: GO FAIR Leiden hosts the Centre of the Hourglass alignment meeting.
- September 9-10, 2019: GO FAIR Leiden visits US National Academy of Sciences, Engineering, and Medicine meeting in Washington DC: joint workshop on the FAIR Digital Object Framework.
- September 19-20, 2019: CODATA 2019: Towards next-generation data-driven science: policies, practices and platforms, Beijing: Multiple sessions covering FAIR Digital Object Framework.
- October 22, 2019: 14th RDA Plenary Meeting in Finland, Helsinki: A co-located event on the FAIR Digital Object was held by GEDE Working Group.
- October 28-29, 2019: Paris Workshop on the FAIR Digital Object.
- January 1, 2020: Joint Statement on FAIR Digital Object Framework is published and is open for participation.
- January 21-22, 2020: FAIR Molecule Hackathon, Hamburg: creating a FAIR Digital Object for molecular structure (slides).
- January 23-24, 2020: Annual meeting of GO FAIR Implementation Networks, Hamburg: Accelerating Convergence in 2020.