Workflow for building interoperable food and nutrition security (FNS) data platforms

Background: In response to growing needs for the integration of heterogeneous data on food and nutrition se-curity (FNS), and the current fragmentation of interoperability resources, the 'FNS-Cloud project' aims to develop a cross-domain, interoperable data platform that integrates diverse FNS data. Currently, there is insufficient guidance on how to develop such an FNS data platform and integrate a variety of FNS data types that differ in both their syntax and semantics.Scope and approach: In the present study, we propose a generalizable workflow to guide data managers in building interoperable, cross-domain FNS data platforms, which centres around the definition of interoperability criteria that capture standardized data structures, terminologies and reporting formats for key variables across FNS data types. Information technology tools for automating different workflow steps are discussed. Finally, we include an illustrative case study, where we harmonize and link branded food datasets based on pre-defined interoperability criteria to answer an example research question.Key findings and conclusions: Our work highlights the unique harmonization requirements within the FNS field. We provide two examples of how generic and domain-specific interoperability criteria addressing these re-quirements can be defined. Incoming FNS data must comply with defined criteria in order to enable their (semi-) automated integration into any data platform. Our case study reinforces the importance of semantic annotation of FNS data, and the need for clear mapping rules to be included into platform-internal semantic data models. The proposed workflow can be applied to any setting in which data managers strive towards harmonized and linked FNS data, and, thus, promotes an open-data and open-science environment.