Open nickynicolson opened 5 months ago
Hi @nickynicolson.
Thanks for raising this. While we can definitely look into improving the API and/or csv export, the general usefulness of such changes might be limited. Perhaps I can help you with a custom export to get you what you're interested in?
I'm interested in determining how many cited uses of GBIF mobilised data are to dataset DOIs, in comparison to download DOIs. The data returned from the literature export API endpoint is insufficient to answer this query. Example below compares the data available from the literature search API (1) with the tabular export endpoint (2). The former includes a reference to the dataset 10.15468/ab3s5x, not included in the latter. Users are directed to use the export through the "download as TSV" link shown on the resources page - this suggests that it will include all information used to format the data displayed on that page. Without the details for dataset level citations, this is not the case.
The contents of the tag
gbifDOI
from the JSON API results are shown in the tabular data columngbif_download_key
when the doi represents a download (with multiple values separated by the pipe character). Can a similar column (e.g.gbif_dataset_key
) be added containing the ID(s) of the referenced dataset(s) when the DOI represents a GBIF dataset? (Note - not looked closely into this, but might want to investigate similar issue with the derived dataset DOIs - see eg: https://api.gbif.org/v1/literature/21b2480f-f641-3188-980e-536f130cea01)(1) Results from https://api.gbif.org/v1/literature/search?q=ecophysiology+green+crab
(2) Tabular format record extracted from https://api.gbif.org/v1/literature/export?format=TSV