gbif / hp-uk-collections

Source for the DiSSCoUK Data Portal provided by GBIF Hosted Portals
https://dissco-uk.org/
2 stars 1 forks source link

publishingCountry occurrence filter includes institutions outside UK #4

Open lmfrench opened 2 years ago

lmfrench commented 2 years ago

Occurrences from institutions outside the UK can be found on the portal in the specimen search.

e.g. this MNHN specimen: https://uk-collections.hp.gbif-staging.org/specimen/search?entity=3864002911&institutionKey=6a6ac6c5-1b8a-48db-91a2-f8661274ff80

This is included in our portal because it is part of this dataset: https://www.gbif.org/dataset/2ecef267-0e17-4f3f-adf9-20b036e6167a , which is published by SCAR - AntOBIS (an institution in the UK) https://www.gbif.org/publisher/104e9c96-791b-4f14-978c-f581cb214912

More examples can be found looking at this api query: http://api.gbif.org/v1/occurrence/search?basisOfRecord=PRESERVED_SPECIMEN&publishing_country=GB&limit=0&facet=institutionCode&facetLimit=600

MortenHofft commented 2 years ago

Hmm. Yes as mentioned when I configured that example filter, it isn't ideal unfortunately. I think of the institution, collection and specimen scope more as a starting point for you to work with.

But before putting effort into the specimens scope, it might be worth thinking about what institutions and collections to include. An how those are filtered. When we know that, then I think we are in a better position to find a filter for the specimens (which might include adding an additional filter to the occurrence API or might be as simple as providing a list of institutions).

There are currently 643 institutions and 580 collections on your site. But you mentioned that you expected 100 institutions.

I mentioned that because 100 institutions might be reasonable to add as a filter on specimens, but 640 on the other hand might be a performance killer. It could also be that we added an additional field to occurrence search. E.g. institutionCountry, but it really depends on what we have in GrSciColl and what you want to show.

Let me know if I can be of any help

lmfrench commented 2 years ago

Thanks - is institutionCountry from GrSciColl? If so, we can make sure all institutions we want to show have this included in their grscicoll record.

For now this issue is on hold, we need to decide how we want to launch the portal first (we might want a limited list of institutions to start with, and if we did this any filter we applied now to fix this issue would need to be changed)

MortenHofft commented 2 years ago

is institutionCountry from GrSciColl

Currently there is not such filter as institutionCountry on occurrences. It was just a loose idea/example for a filter we might be able to add to the occurrence API if that was the best way to define the scope.

Anyway, I'm fairly confident we can find a solution somehow. Ping me when you are ready