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.

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

Dataset publications [RDSP] #63

Closed
jpullmann opened this issue Jan 18, 2018 · 8 comments

Comments

8 participants
@jpullmann
Copy link
Contributor

commented Jan 18, 2018

Dataset publications [RDSP]

Provide a way to link publications about a dataset to the dataset.


Related requirements: Profiles listing [RPFL] Dataset aspects [RDSAT] Dataset publications [RDSP] 
Related use cases: Common requirements for scientific data [ID9] 
@nicholascar

This comment has been minimized.

Copy link
Contributor

commented Jan 31, 2018

Is this an inverse of a prov:wasDerivedFrom where a Publication is wasDerivedFrom the Dataset?

PROV recommends the inverse prov:hadDerivation but doesn't declare it.

@andrea-perego

This comment has been minimized.

Copy link
Contributor

commented Jan 31, 2018

Is this an inverse of a prov:wasDerivedFrom where a Publication is wasDerivedFrom the Dataset?

PROV recommends the inverse prov:hadDerivation but doesn't declare it.

In the JRC Data Catalogue we actually use dct:isReferencedBy, that [PROV-DC] defines as a subproperty of prov:hadDerivation:

https://www.w3.org/TR/prov-dc/#term_isReferencedBy

@nicholascar

This comment has been minimized.

Copy link
Contributor

commented Feb 2, 2018

Interesting. To my reading, isReferencedBy is a looser association than hadDerivation. The former indicates a only somewhat influential association (A isReferencedBy B means B knows about A but isn't wholly resulting from actions on A) whereas the latter indicated a much closer association (A hadDerivation B means B comes from A in some substantive way in which work was done on A to produce B).

If we are talking a publication that is about a dataset then the publication is a result of work on on the dataset and the relation is close, not a passing reference. I would choose hadDerivation.

@andrea-perego

This comment has been minimized.

Copy link
Contributor

commented Feb 3, 2018

Interesting. To my reading, isReferencedBy is a looser association than hadDerivation. The former indicates a only somewhat influential association (A isReferencedBy B means B knows about A but isn't wholly resulting from actions on A) whereas the latter indicated a much closer association (A hadDerivation B means B comes from A in some substantive way in which work was done on A to produce B).

Looking at the mappings defined in [PROV-DC], it seems that, in the intention of the PROV WG, prov:hadDerivation is more generic, as it is defined as superproperty of both dct:isReferencedBy and dct:hasFormat - thus denoting different types of derivation.

If we are talking a publication that is about a dataset then the publication is a result of work on on the dataset and the relation is close, not a passing reference. I would choose hadDerivation.

This is one of the cases, but you may have a publication just referring to a dataset as related work, or documenting how the dataset has been created. Moreover, it is often the case that a publication has different types of relationship with a dataset.

For instance, our requirement was to provide the list of publications about a given dataset - so, those "citing" a given dataset, which is one of the cases included in the definition of dct:isReferencedBy:

A related resource that references, cites, or otherwise points to the described resource.

However, if a publication is documenting how a dataset has been created, we specify it both with dct:isReferencedBy and dct:provenance:

a:Dataset a dcat:Dataset ;
  dct:isReferencedBy a:Publication ;
  dct:provenance a:Publication .

a:Publication a foaf:Document ;
  ...
@agbeltran

This comment has been minimized.

Copy link
Member

commented Jan 9, 2019

When considering linking datasets and publications or other references, the HCLS Dataset Description community profile recommends:

  • dct:references for web pages and documents
  • cito:citesAsAuthority to include publications about the dataset
  • rdfs:seeAlso for other references
@agbeltran

This comment has been minimized.

Copy link
Member

commented Jan 9, 2019

For DATS, we found the requirement to make the distinction between primary publications (one or more publications that describe how the dataset was produced) and other references or citations (publications that refer / cite or use the dataset):

https://github.com/datatagsuite/schema/blob/29134fd4aa16f7e76b6b9e20238c7b4de82b8518/dataset_schema.json#L112

For annotating these properties, we used sdo:citation in both cases, as schema.org doesn't distinguish between primary or other publications:
https://github.com/datatagsuite/context/blob/master/sdo/dataset_sdo_context.jsonld#L29
Same for context files based on OBO ontologies, we relied on iao:citation in both cases:
https://github.com/datatagsuite/context/blob/1ea30cbda28c814ed3635994264d679c3af4eca9/obo/dataset_obo_context.jsonld#L18

Given the introduction of dcat:Resource, this requirement should apply to any catalogued resource (including datasets and services).

@agbeltran

This comment has been minimized.

Copy link
Member

commented Mar 6, 2019

@andrea-perego:

However, if a publication is documenting how a dataset has been created, we specify it both with dct:isReferencedBy and dct:provenance:

a:Dataset a dcat:Dataset ;
  dct:isReferencedBy a:Publication ;
  dct:provenance a:Publication .

a:Publication a foaf:Document ;
  ...

The range for dct:provenance is dct:ProvenanceStatement, so could it be a foaf:Document?

@riccardoAlbertoni

This comment has been minimized.

Copy link
Collaborator

commented Mar 6, 2019

@agbeltran wrote:

When considering linking datasets and publications or other references, the HCLS Dataset Description community profile recommends:

  • dct:references for web pages and documents
  • cito:citesAsAuthority to include publications about the dataset
  • dfs:seeAlso for other references

I haven't found the actual axiom or explanation saying that, but I guess dct:references and dct:isReferencedBy are one the inverse of the other, so +1 to use

a:Dataset a dcat:Dataset ;
  dct:isReferencedBy a:Publication .

a:Publication a foaf:Document .

The range for dct:provenace is dct:ProvenanceStatement, so could it be a foaf:Document?

I might be wrong, but I don't see much restriction about dct:ProvenanceStatement, it seems just a class, which potentially can be anything.

DCAT revision automation moved this from To do to Done May 21, 2019

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.