Towards FAIR Scientific Notebooks
Nearly all of modern scientific inquiry is built upon data and analyses that are the result of complex computational workflows. Large scale scientific software packages are the glue of much of this inquiry, but increasingly, tools such as Jupyter Notebooks, Beaker, Zeppelin and others, are supporting scientific inquiry through open, interactive and executable containers of scientific notebooks. The growth in data and scientific software, alongside these scientific notebooks, is exposing challenges for facilitating discovery and access to all of these important artifacts.
FAIR Data Principles [Wilkinson, et al, 2016] have been proposed as guidelines to facilitate discovery of data, but these same principles are also intended to apply to "all scholarly digital research objects ... to ensure transparency, reproducibility, and reusability." These guidelines can be applied to the proliferation of notebooks that are becoming crucial to all modern scientific workflows, and perhaps it is time to take the future needs of our scientific communities more seriously and consider how to integrate and apply FAIR principles.
In this poster, we propose that we develop, adopt and accelerate FAIR principles for scientific notebooks by
- raising community awareness of the benefits of FAIR,
- developing community norms that promote FAIR principles, and
- supporting the development of standards and tools that promote FAIR best practices.
If we do not move toward FAIR principles for scientific notebooks, we put important data, analyses and intermediate outcomes that will form the backbone of tomorrow's scientific breakthroughs at risk of becoming unusable.
NOTE: This poster was presented as part of the Scientific Sotware Days (SciSoftDays.org) 9th Annual Workshop, April 26-27, 2018 at the University of Texas, Austin.
