How can we extract information from NIH Reporter to populate an initial version of a DSP for a particular project. Can we use the LLM approaches to curation?
Ideal for this to be coupled with grant curation process, but this could be done with existing grants in the CCKP.
This seems like low hanging fruit to give us a sense of expected data types (even if high level - e.g. species level), to give us a basis for initial conversations for contributors (reduces the burden on them to populate a DSP from a blank template.
Note:
do not want this we be a blocker to proceeding with the grant curation outlined in https://github.com/mc2-center/data-models/issues/145. If this is not deemed trivial, pursue grant curation and design this further after the fact.
May be combined with https://github.com/mc2-center/data-models/issues/145 or designed in parallel with this work.
Consider:
Note: