Closed KonradHoeffner closed 4 years ago
SELECT ?cit COUNT(?swp)
{
?cit a [rdfs:subClassOf hito:Citation].
OPTIONAL
{
?swp a hito:SoftwareProduct; ?p ?cit.
}
}
GROUP BY ?cit
HAVING (COUNT(?swp)>1)
ORDER BY DESC(COUNT(?swp))
This occurs due to the URI generation out of the label in CSV2RDF. A randomly generated ID of a large length would have a very high probability of solving the problem but make the URIs completely unreadable, which I don't like for Semantic Web resources. Thus, I will first prefix the citation URIs generated in CSV2RDF with the name of the software product and the type of the citation.
For example, hito:Reporting could become split into:
Result after CSV2RDF prefixing
As they don't have any prefix, they seem to be manually specified. Manually prefix products.ttl and check again afterwards.
The problem is that the citations are defined in hito.ttl and reused, which is not what is intended now.
Forbidden due to cardinality rules. Related to #39.