Process tables individually (analysis level)

[jdkent]

We want to be able to process the tables in the context of the papers to separate out coordinates into meaningful analyses.

[jdkent]

from alejandro commenting on extraction stage:

I think it's just impossible to make sense of without opening up the paper. Meta data is also sometimes better than others (i.e. Table number, vs a random name for Analysis)