D6.3 Top 10 FAIR Data Things in Artificial Intelligence & Published recommendations on FAIR data in Health Technology

This is a deliverable of two parts: a “Top 10 FAIR Data Things” in Artificial Intelligence and FAIR Data for Health Technology (HT). While there have been efforts to implement the FAIR principles in specific fields applying Artificial Intelligence (AI), there is a dearth of clear low-threshold, cross-discipline guidelines. FAIR practices, could benefit AI research and researchers e.g., through setting benchmarks for model development and interpretation, enabling reuse in research and education and leading to better collaboration. Using the Delphi Technique we engaged the AI research community to create a “Top 10 FAIR Data Things in AI”, intended to be a simple- to- follow guide and the foundation for more nuanced strategies as needs arise. Regarding FAIR practices in HT, there is no ready community, and negligible literature. We surveyed researchers and data professionals to obtain a snapshot of current awareness and practice regarding the FAIR principles. Here we present a preliminary discussion on the issues, intended to stimulate discussions in other areas of HT research, supporting FAIR data practices more widely across this domain and the focus of a relevant network being initiated as part of this task.