suggest to index Plazi's treatment bank

[jhpoelen]

hi @dimus et al.

Are you still maintaining globalnames.org ?

If so, I was hoping you can consider the following:

Plazi https://plazi.org keeps an extensive list of taxonomic literature and associated taxonomic names.

Plazi exports these taxonomic literature <> name links as DwC-A and register them with GBIF.

You can find their publications at https://www.gbif.org/occurrence/search?dataset_key=6384b520-7e9f-4874-a414-76c2e9b01d74&type_status=TYPE .

As a user (GloBI), I would like to be able use the Global Names resolvers to find taxonomic treatments in Plazi.

The taxonomic treatments can be located by linking the TaxonId fields that are available in the Plazi publications.

for example, when I lookup: Rhinolophus denti , I expect to find a Plazi name with id 885887A2FFC88A21F8B1FA48FB92DD65.taxon (also see https://www.gbif.org/occurrence/2597533915). This identifier can then be translated into a link to the related taxonomic treatment via http://treatment.plazi.org/id/885887A2FFC88A21F8B1FA48FB92DD65 .

See also attached screenshots.

fyi @myrmoteras

[dimus]

@jhpoelen good point, noted, put it on a back burner. We got some troubles with Scala version of resolver, moving it to Go atm. I will add Plazi when resoltuion is migrated to new code (a few months from now).

[jhpoelen]

@dimus Thanks for considering indexing the Plazi taxonomic names associated to the many treatments they index.

Is there a way I can run my own resolver locally with my own indexed names? What would it take to replicate your setup? I'd actually prefer running tools in a controlled environment anyway, because of the variability of web services (e.g., limited network bandwidth, single point of access).

[dimus]

Is there a way I can run my own resolver locally with my own indexed names? What would it take to replicate your setup? I'd actually prefer running tools in a controlled environment anyway, because of the variability of web services (e.g., limited network bandwidth, single point of access).

I do want to make resolver "localizable". New one is located at https://github.com/gnames/gnames/. The idea i have it so include docker-compose file that sets up all components locally. I do not have a decent harvester yet sadly, so thats gives a choice of either building DB out of a dump or populating DB by a script.

The project is not finished yet, but I do hope it to be running by October. It would be awesome if you help to test localization when I get to a point of a 'rea' release!

[jhpoelen]

@dimus excellent! I'd be happy to help test your tool, especially if the features align with those of Nomer:

1. easy to install (in my opinion, docker is not easy for most and usually just hides overly complex infrastructures)
2. able to index, cache and version existing taxon lists (e.g., index Plazi names from taxon lists published in DwC-A)
3. able to batch export / stream link results
4. command-line interface
5. able to run on single laptop
6. designed for offline (no internet) workflows
7. separate commands to do common name operations: find, parse, resolve

Is this in line with the features you are planning to support?

[dimus]

Hm, looks like there is no aggregated file with all PLAZI data at GBIF. @jhpoelen, any ideas if it exists?

[myrmoteras]

@dimus This is the summary and access to all the datasets in GBIF: https://www.gbif.org/publisher/7ce8aef0-9e92-11dc-8738-b8a03c50a862/metrics
You can also get it from here: http://tb.plazi.org/GgServer/dioStats/stats?outputFields=doc.articleUuid+doc.doi+doc.gbifId&groupingFields=doc.articleUuid+doc.doi+doc.gbifId&limit=100&FP-doc.gbifId=1-&format=HTML
just remove the limit=100 and you get all the 31.115 articles/datasets on GBIF

[dimus]

@myrmoteras thank you very much, let me try, I just sent an email to you asking about this, so you do not need to answer it :)

[myrmoteras]

are you aware of the stats in Plazi? http://plazi.org/api-tools/statistics/
There are two that allow you qurey the article and treatment stats.
For example you could get all the verbatim taxon name of the treatment, the link to the article, the treatment (eg http://tb.plazi.org/GgServer/srsStats/stats?outputFields=doc.uuid+doc.articleUuid+tax.name&groupingFields=doc.uuid+doc.articleUuid+tax.name&limit=100&format=HTML) and much more

[dimus]

@myrmoteras, I found this link at http://plazi.org/api-tools/statistics/:

http://tb.plazi.org/GgServer/xml.rss.xml

with items like:

<item>
<title>Caccothryptus arakawae Matsumoto 2021, sp. nov.</title>
<description>Caccothryptus arakawae Matsumoto 2021, sp. nov. (pages 171-171) in Matsumoto, Keita 2021, Six new species of the genus Caccothryptus from the Himalayas (Coleoptera: Limnichidae), European Journal of Taxonomy 739, pages 168-184</description>
<link>http://tb.plazi.org/GgServer/xml/F035D85E9906FFF68106FBE4FB925511</link>
<pubDate>2021-03-19T09:23:16-02:00</pubDate>
<guid isPermaLink="false">F035D85E9906FFF68106FBE4FB925511.xml</guid>
</item>

```

I can probably use this file, currently it has 418765 items, which sounds about right

[jhpoelen]

For nomer, I've been using https://github.com/plazi/treatments-rdf . Versioned and all!

[dimus]

Thanks @jhpoelen! I already impored PLAZI (data source #194) with less metadata to https://resolver.globalnames.org/data_sources/194, and will transfer these data to https://verifier.globalnames.org in the evening.

I am going to close this ticket and make a new one about using your approach, so we have more information about names.

[dimus]

@myrmoteras, @jhpoelen the first version of harvesting will propagate to https://verifier.globalnames.org in a couple of hours.

And I made #47 to remember to enhance the harvest in the future according to @jhpoelen suggestion.