Schema for figure data

[blaiszik]

Figures have a very specific shape, and it may be helpful to capture this in a schema block

For example, fig description, axis labels, units, etc.

[jgaff]

Are we going to try and separate image from figure, then? Also, how will the figure information be populated?

[blaiszik]

For now, I envision this being populated manually (i.e., putting a JSON formatted file in the same directory as the figure). This came up specifically in the context of the NIST/CHiMaD Polymer data special issue call. Much of the data is essentially the figures for a given paper, this would provide a nice way to store the high level info about them and make them searchable.

[jgaff]

So are we separating image and figure? If the figure is an image it is parsed as an image right now. Will we need to check somehow if the image is a figure, and only populate the image block for non-figures?

It seems to me, though, that the better solution to "make figures searchable" is for the researchers to provide the data for the figures instead of indexing the figures themselves.