Not well-formed IRI in bio2rdf.org dataset

[karlisc]

Dear bio2rdf.org maintainers,

while trying to understand the bio2rdf.org data structure, I find out that there is 21 triple listing
http:purl.org/dc/dcmitype/Dataset
as the object in rdf:type assertions, which is not a well-formed IRI, of course.

Can this be corrected, or is this meant to stay and to be handled in some ways by data set clients?

Thanks!
Best regards,
Kārlis

[karlisc]

This is easy to find by the query
select ?c (count(?x) as ?cx) where {?x a ?c} group by ?c order by desc(?cx)
and then scrolling to the count 21 in the result list.

I have found that by trying to examine the data set structure for the purpose of offering visual schema diagrams for the data set fragments and context-sensitive SPARQL query auto-completion both in ViziQuer tool (https://viziquer.lumii.lv) and possibly outside it, as well.

[micheldumontier]

well, that comes from the disgenet dataset, not a bio2rdf one. just recommend to adjust accordingly.