Re:A Reboot of the Legendary Physics Site ArXiv Could Shape ...
by westurner on Thursday May 12, 2016 @06:33PM
(#11687338)
> a way to evaluate controls
a way to evaluate premises (assumptions, controls, data, data transformations) and conclusions (presented as e.g. JSONLD, RDFa in a standard form (potentially like IPython/Jupyter .ipynb; but with an OrderedMap of I/O sequences with fixed #urifragment IDs)
https://twitter.com/search?q=%23LinkedReproducibility
https://twitter.com/search?q=%23MetaResearch
- schema.org/MedicalTrialDesign enumerations could/should be extended to all of science (and then added to all of these PDFs without structured edge types like e.g. {intendedToReproduce, seemsToReproduce} (which then have specific ensuing discussions))
- http://health-lifesci.schema.org/MedicalTrialDesign
- there should be a way to evaluate controls in a structured, blinded, meta-analytic way
- PDF is pretty, but does not support RDFa (because this is a graph)
... notes here: https://wrdrd.com/docs/consulting/data-science#linked-reprod...
(edit) please feel free to implement any of these ideas (e.g. CC0)