edm:hasView is output in addition to edm:isShownAt

[kba]

Might be a bug in the super-property logic in Dm2e2Edm

[kba]

The problem here is that the uber/dingler data contain patterns like

?cho edm:isShownAt ?uri .
?cho dm2e:hasAnnotatableVersionAt ?uri .

with ?uri being identical for both statements.

Now, the DM2E model defines dm2e:hasAnnotatableVersionAt as

dm2e:hasAnnotatableVersionAt a owl:ObjectProperty ;
    rdfs:subPropertyOf edm:hasView .

Hence, the DM2E-EDM converter notices that dm2e:hasAnnotatableVersionAt is not part of EDM but edm:hasView is and will therefore output the pattern

?cho edm:isShownAt ?uri .
?cho edm:hasView ?uri.

Unless the DM2E model skips subclassing dm2e:hasAnnotatableVersionAt to edm:hasView or subclasses it to edm:isShownAt (which would remove the redundancy if both dm2e:hasAnnotatableVersion and edm:isShownAt refer to the same URI, I'd say, this is a WONTFIX.

[JIwanowa]

I think the best solution would be to skip any subclassing of dm2e:hasAnnotatableVersionAt.

[d0rg0ld]

I actually have a different proposal: hasAnnotatableVersion should be very close to "isShownBy". So I'd suggest to convert hasAnotatableVersionAt to isShownBy whenever there is no isShownBy Property assigned right away.

[d0rg0ld]

otherwise skip of course

[JIwanowa]

In case of dingler the URIs for dm2e:hasAnnotatableVersionAt and edm:isShownAt are identical. Therefore it wouldn't make sense to convert dm2e:hasAnnotatableVersionAt to edm:isShownBy and to deliver the duplicates to Europeana. Furthermore we have additional URIs for edm:isShownBy which leads to the images. Apart of that the definition of edm:isShownAt allows only URIs with context information.

[d0rg0ld]

I think that the Dingler case is rather special in this regard, as you link to the "isShownAt" html representation for annotation, which was also noted by Net7 in #47

[d0rg0ld]

I had a brief look at the data and a chat with Julia. In most cases, dm2e:hasAnnotatableVersionAt is assigned to page level CHOs and should therefore be no issue for Europeana ingestion. For the few cases where hasAnnotatableVersion is assigned to objects on item level, it cant be clearly distinguished whether they should be assigned to isShownAt or isShownBy. Therefore I now also propose to skip the subclassing of hasAnnotatableVersionAt ...