If I add a class in the DBpedia ontology as an equivalent class to a Wikidata class (
P1709), is the reverse then automatically picked up by DBpedia as well? Or should I still add the Wikidata equivalent class separately to the DBpedia mappings wiki?
I am also wondering about this. I am not aware of any process that would automatically do this. Please correct me if I am wrong. To be sure, I suggest you add the equivalent class or property to both ontologies.
You can add equivalentClasses (DBpedia → Wikidata) in the mappings wiki, see here: OntologyClass:Person - DBpedia Mappings
Recently, we did some analysis of Wikidata and the data is quite hard to use, see e.g. here Giving knowledge back to Wikipedia: Towards a Systematic Approach to Sync Factual Data across Wikipedia, Wikidata and External Data Sources - DBpedia Association or here https://svn.aksw.org/papers/2020/qurator_gfs/public.pdf . So Wikidata seems (1) quite behind Wikipedia in recency, (2) the links are very few and also have unclear semantics, i.e. they use properties with undefined semantics, which messes up our global clusters (3) there is no formal meaning for P1709. Then there is the fact that there is a lot of good linked data out there, e.g. the DNB or others that you can aggregate easily (or we intend to make it more findable and thus easier) and Wikidata takes the data from there anyhow with non-systematic bots.
So our tendency is to use less Wikidata, simply because it doesn’t work well for us quality-wise. The lack of a Wikidata ontology also aggravates that. dbo:Person also has two equivalent class links to Wikidata, because they have person, but it says to use human
(avoid use with P31; use Q5 for humans) ??
Together with Archivo, we will soon start to link all owl ontologies better and use the DBpedia Ontology as a backbone for this. There will be a discussion in the DBpedia Ontology session at SEMANTiCS where this is headed: DBpedia Day @ Semantics 2021 - DBpedia Association