Modeling community standards for metadata as templates makes data FAIR

Sharing experimental data and making data findable, accessible, interoperable, and reusable (FAIR) are of growing importance to funders, publishers, and to scientists themselves. Evaluating whether datasets are FAIR is challenging because the FAIR Guiding Principles include highly idiosyncratic criteria regarding the metadata used to annotate datasets -- such as whether metadata are "rich" and whether they adhere to "domain-relevant" community standards. Rather than attempting to assess whether the metadata comport with the FAIR principles after datasets are already archived in some repository, an alternative approach is to enable scientific communities to define their own templates for metadata that necessarily encode these discipline-specific elements. Such templates, inherently, incorporate the level of "richness" desired by the community and the standards (both reporting guidelines and ontologies) that the community wishes to adopt. We have explored this approach in the context of two software systems that use such metadata templates. One is the CEDAR Workbench, which is used to author both metadata templates and new metadata. The other is the FAIRware Workbench, which evaluates the metadata of archived datasets for their adherence to community standards. There are benefits when templates for metadata are made to be central elements of an ecosystem of tools to manage online datasets -- both because they serve as an examinable community reference for what constitutes FAIR data, and because they embody that perspective in a machine-actionable format that can be distributed among a variety of software applications to assist with data stewardship and data sharing.