I've been meaning to demonstrate that doing what @cmungall suggested in https://github.com/cthoyt/orcidio/issues/5 was feasible. This PR consists of about a half hour of find/replace in the ro-edit.owl file in a text editor where I replaced string mentions of people's names in provenance fields with canonical ORCID identifiers as URIs. I put a decent debt in this repository's provenance debt (i.e., turning unstructured provenance into structured provenance).
If this PR is accepted, here are some future to-do's:
Continue updating the remaining instances dc, dcterms, dce, and iao annotations that point to individuals
Import ORCIDIO so all of these beautiful, clean ORCID identifiers get names overlayed when editing in Protege
After finishing 1) to completion, add additional unit tests to make sure nobody ever adds unstructured information again (@matentzn I hope you could help introduce this one predicate at a time, perhaps starting with IAO_0000117 since this is fully cleaned up now)
I've been meaning to demonstrate that doing what @cmungall suggested in https://github.com/cthoyt/orcidio/issues/5 was feasible. This PR consists of about a half hour of find/replace in the ro-edit.owl file in a text editor where I replaced string mentions of people's names in provenance fields with canonical ORCID identifiers as URIs. I put a decent debt in this repository's provenance debt (i.e., turning unstructured provenance into structured provenance).
If this PR is accepted, here are some future to-do's:
dc
,dcterms
,dce
, andiao
annotations that point to individualsIAO_0000117
since this is fully cleaned up now)