OK. Back to integrating the Zenodo API onto Open Context.
Another question for the #beforemoderntimes #DH #openscience #opendata set.
Shall I use EZID to mint ARK identifiers for all 1.5 million URIs managed by OC? @paregorios will Pleiades plan on minting persistent IDs for all Pleiades entities or you happy with DOIs for data dumps, and not bother with persistent IDs for more granular data?
I'm just trying to judge what is a good use of my time, and what's broadly useful.
(cc @steko)
No. I guess we're all trying to guess what the future will bring. I'm a little worried about identifier soup, but I want people to be able to formally cite a specific record. It seems then that some sort of persistent ID scheme, backed by institutions beyond our own, is needed?
I like ARKs. Cheap, easy to mint. PeriodO uses them.
Now, thanks to @ekansa, I'm #reading:
“Archival Resource Key (ARK) Identifiers.” N2T.net: Resolver: Names -> Things. Accessed May 2, 2018. http://n2t.net/e/ark_ids.html.
Zotero: https://www.zotero.org/users/465/items/J9IHLSUY
To get up to speed on #ARKs so I can consider better whether we should start using them for resources in @pleiades in addition to the #CoolURIs we already mint ourselves.
Yet another rabbit hole to fall into.
John Kunze with CDL is a key architect of NOIDs, and he has lots to say about identifiers. Like everything else important, there is much to this subject.
@ekansa @paregorios what does a typical OC URI look like? I think "cool URIs" don't need an additional layer of persistence, unless 1) you've got a separate archival resource, perhaps for long term preservation 2) the additional PID is transparent, e.g. it has a different prefix than the HTTP base URL, but keeps the same suffix identifier.
Makes sense?
Here's a typical URI:
https://opencontext.org/media/48fb6172-647a-40f6-a576-e50d006a4f52
I also have an ARK id for it:
https://n2t.net/ark:/28722/k2q244j7j
The JSON-LD representation (https://opencontext.org/media/48fb6172-647a-40f6-a576-e50d006a4f52.json) says "owl:sameAs" for the ARK id.
I just want to maximize the probability that someone can look up the citation to an OC record over the years.
@steko @paregorios
@aejolene
@sebhth
What do you all think? Will it help, harm, or "meh" to add ARK identifiers to all our URI identified data?
@paregorios @steko @aejolene @sebhth
The effort is minimal, because I can script it. Getting ARKs from EZID just requires API calls.
So, ok, I'll get on it.
@ekansa @paregorios @aejolene @sebhth so ARKs are free to mint? If so, why is #DOI favoured?
@paregorios @aejolene @sebhth @steko
I think Tom called it. Also DOIs get more indexing in different discovery and metrics (including Altmetrics).
EZID supports minting 1 million ARKs / year at $1500 / year service fee.
@ekansa @paregorios @aejolene @sebhth btw, Altmetrics is part of Digital Science which is part of the Holtzbrinck Publishing Group https://www.holtzbrinck.com/
@ekansa
Since @pleiades mints cool https URIs for our place, name, location, and connection resources, I'm knee-jerk disinclined to create multiple IDs for the same things in other namespaces. Does that make me a bad person?
@steko