Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

OAI-PMH => ORE? #149

Closed
hrzepa opened this issue Apr 13, 2021 · 18 comments
Closed

OAI-PMH => ORE? #149

hrzepa opened this issue Apr 13, 2021 · 18 comments

Comments

@hrzepa
Copy link

hrzepa commented Apr 13, 2021

We use ORE for harvesting rather than OAI-PMH. Can you detect this?

Eg https://doi.org/10.14469/hpc/6215 or https://doi.org/10.14469/hpc/6880
@huberrob
Copy link
Contributor

F-UJI apparently has problems with the DOIs you provided, but the reason seems some re3data entries F-UJI is trying to use.. I'll try to fix this

Regarding OAI-ORE, F-UJI is not yet able to handle this, can you please provide a endpoint URI which provides the ORE resource map file?

@hrzepa
Copy link
Author

hrzepa commented Apr 13, 2021 via email

@huberrob
Copy link
Contributor

Rearding ORE, it would be not a big problem to implement this in F-UJI.
The problem for F-UJI with the given examples is that a link between the PID landing page and the ORE file seems to be missing.
According to the specification (http://www.openarchives.org/ore/1.0/http) content negotiation should be possible (e.g. Accept: application/atom+xml) but this seems not to work on your site. Alternatively, links to ORE could also be expressed in the head section of the landing page like this:

But I could not find it in the example data sets you provided. Could you please check?

Regarding the data standards, I think F-UJI could detect these formats, but only if there would be the possibility to find ORE files from the PID landing page as starting point (see above).

I would love to learn more about the IUPAC approach re metadata standardisatio. I think many communities now switch to more generic formats like schema.org or DCAT which are quite flexible.

@hrzepa
Copy link
Author

hrzepa commented Apr 13, 2021 via email

@huberrob
Copy link
Contributor

It is trivial to insert and I will ask the programmers if they could do that.

Great! can you please let me know once this is implemented? I will try to integrate ORE support immediately once this is ready..

@hrzepa
Copy link
Author

hrzepa commented Apr 15, 2021 via email

@huberrob
Copy link
Contributor

huberrob commented Apr 15, 2021 via email •
edited
Loading

@hrzepa
Copy link
Author

hrzepa commented Apr 28, 2021 via email

@huberrob
Copy link
Contributor

huberrob commented May 3, 2021

I'll try to answer your questions below:

  1. FsF-F3-01M - Metadata includes the identifier of the data it describes. The metadata includes eg

10.14469/HPC/6216

Actually, F-UJI tests the presence of data links here, not the inclusion of the DOI or so in metadata (this is done in FsF-F2-01M). The output for FsF-F3-01M is a bit confusing..
I think F-UJI will recognize links to data in the OAI-ORE resourcemap.

  1. FsF-A1-01M - Metadata contains access level and access conditions of the data and
    FsF-A1-03D - Data is accessible through a standardized communication protocol.

Can I presume that when eg https://data.hpc.imperial.ac.uk/resolve/?ore=6216

is elevated to

For FsF-A1-01M access levels such as 'closed' 'public' have to be defined such as http://vocabularies.coar-repositories.org/documentation/access_rights/ The resourcemap will not solve this problem since these rights have to be explicitely declared in the metadata.
But FsF-A1-03D will be passed using the resourcemap info in case F-UJI finds links to data files there.

  1. Metadata specifies the content of the data.

I presume once

is declared, this will retrieve eg the media type of the data, which is declared at
https://data.hpc.imperial.ac.uk/resolve/?ore=6216

Yes, F-UJI will identify media types and file sizes in the resourcemap which is sufficient to verify the data content.

  1. FsF-R1.3-02D - Data is available in a file format recommended by the target research community.

Will this also be addressed by item 4 (declaration of media types). But how does one declare the target research community itself. We use a subject line to do this

The test does not require to specify a community or discipline, F-UJI checks if data formats are community specific (in your case chemistry) or generic scientific file formats (matlab etc..) As far as I can see, the dataset provided will pass the test based on the resourcemap info provided.

I understand that your ICT staff needs some time to embed the resource map link in the landing page, no problem, just let me know once this is ready..

best regads,
Robert

@hrzepa
Copy link
Author

hrzepa commented May 3, 2021 via email

@huberrob
Copy link
Contributor

huberrob commented May 3, 2021

Dear Henry,

You can also use other vocabularies:

http://vocabularies.coar-repositories.org/documentation/access_rights/
http://purl.org/coar/access_right
http://purl.org/eprint/accessRights
http://publications.europa.eu/resource/authority/access-right

(some of them unfortunaletly are currently offline)

DataCite schema is very focused on licenses regarding their rightsList element. But According to their specs ("Any rights information for this resource"), it can also be used to define access rights and conditions.

I could imagine something like:

<rightsList>
    <rights rightsURI="http://purl.org/eprint/accessRights/RestrictedAccess" rightsIdentifier="RestrictedAccess" rightsIdentifierScheme="EPRINT" schemeURI="http://purl.org/eprint/">Restricted Access</rights>
  </rightsList>

But maybe @mfenner can explain this in more detail..

<rightsList>
    <rights rightsURI="https://creativecommons.org/publicdomain/zero/1.0/legalcode" rightsIdentifier="cc0-1.0" rightsIdentifierScheme="SPDX" schemeURI="https://spdx.org/licenses/">Creative Commons Zero v1.0 Universal</rights>
  </rightsList>

Is recognized by F-UJI as licence information which is good since this is needed to pass metric FsF-R1.1-01M (Metadata includes license information under which data can be reused.)

Robert

@hrzepa
Copy link
Author

hrzepa commented May 3, 2021 via email

@huberrob
Copy link
Contributor

huberrob commented May 3, 2021 via email

@hrzepa
Copy link
Author

hrzepa commented May 3, 2021 via email

@huberrob
Copy link
Contributor

huberrob commented May 4, 2021 via email

@hrzepa
Copy link
Author

hrzepa commented May 4, 2021 via email

@huberrob huberrob mentioned this issue May 12, 2021
Open
@hvdsomp
Copy link

hvdsomp commented May 12, 2021

I just discovered this issue by means of a link in #65. A few remarks:

  • Discovery of OAI-ORE Resource Maps is described in the ORE Resource Map Discovery guidelines. Of particular interest, IMO, is the approach that supports discovery of Resource Maps by means of typed links at the level of the landing page of the object described by the Resource Map, using <link> in HTML or a Link header in the HTTP response header, both using the "resourcemap" link relation type, see https://www.openarchives.org/ore/1.0/discovery#ResourceEmbedding. This discovery approach is also illustrated in https://www.openarchives.org/ore/1.0/http#MultiSplash of the ORE HTTP Implementation spec.
  • The content negotiation approach, described in the same section, could be implemented by introducing a new URI (A-1 of the Aggregation in the pic) to which the DOI redirects (you can't implement content negotiation at the DOI because you don't control it). At that new URI one could then implement content negotiation for the landing page (description of the object for humans) or a Resource Map (description of the object for machines).
  • But you're probably not going to go through the effort of introducing such a new URI, changing the DOI redirections, and implementing content negotiation. But you still need a dedicated URI for the Aggregation according to OAI-ORE and the Linked Data principles it is based on. Which I don't see in the Resource Map e.g. https://data.hpc.imperial.ac.uk/resolve/?ore=6216. A simple approach is to append a #aggregation at the end of the Resource Map URI to obtain a dedicated URI for the Aggregation, see https://www.openarchives.org/ore/1.0/http#Simple.
  • Note that the OAI-ORE vocabulary includes the term ore:similarTo to express the relation between the Aggregation and, for example, the DOI (HTTP version) of the object described by the Resource Map associated with the Aggregation, see https://www.openarchives.org/ore/1.0/datamodel#ore:similarTo
  • The spec describing the representation of ResourceMaps in Atom was deprecated quite a while ago, see https://www.openarchives.org/ore/1.0/toc. There is a spec describing a https://www.openarchives.org/ore/0.9/jsonld though.

@hrzepa
Copy link
Author

hrzepa commented May 12, 2021 via email

@huberrob huberrob closed this as completed Mar 4, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@huberrob @hvdsomp @hrzepa