Building a Disciplinary, World‐Wide Data Infrastructure

Sharing scientific data with the objective of making it discoverable, accessible, reusable, and interoperable requires work and presents challenges being faced at the disciplinary level to define in particular how the data should be formatted and described. This paper represents the Proceedings of a session held at SciDataCon 2016 (Denver, 12–13 September 2016). It explores the way a range of disciplines, namely materials science, crystallography, astronomy, earth sciences, humanities and linguistics, get organized at the international level to address those challenges. The disciplinary culture with respect to data sharing, science drivers, organization, lessons learnt and the elements of the data infrastructure which are or could be shared with others are briefly described. Commonalities and differences are assessed. Common key elements for success are identified: data sharing should be science driven; defining the disciplinary part of the interdisciplinary standards is mandatory but challenging; sharing of applications should accompany data sharing. Incentives such as journal and funding agency requirements are also similar. For all, social aspects are more challenging than technological ones. Governance is more diverse, often specific to the discipline organization. Being problem‐driven is also a key factor of success for building bridges to enable interdisciplinary research. Several international data organizations such as CODATA, RDA and WDS can facilitate the establishment of disciplinary interoperability frameworks. As a spin‐off of the session, a RDA Disciplinary Interoperability Interest Group is proposed to bring together representatives across disciplines to better organize and drive the discussion for prioritizing, harmonizing and efficiently articulating disciplinary needs