Open jimkont opened 9 years ago
A problem with providing disjointness axioms is that many of the DBpedia classes are in unusual places in the taxonomy or are not very well defined. For example, one might expect that Canal and Stream should be disjoint but Canal is a subclass of Stream. Classes like Swarm have no effective definition making it very difficult to determine what they should be disjoint from.
You are right, I agree. I'll make a separate issue for improving the class/properties documentation but there is not time to tackle this for the next release. For now we could limit this to top-level classes where the class definitions are usually more obvious
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Disjointness between the top-level concepts is also tricky as there are several whose definition is rather inscrutable.
peter
On 03/05/2015 06:29 AM, Dimitris Kontokostas wrote:
You are right, I agree. I'll make a separate issue for improving the class/properties documentation but there is not time to tackle this for the next release. For now we could limit this to top-level classes where the class definitions are usually more obvious
— Reply to this email directly or view it on GitHub https://github.com/dbpedia/ontology-tracker/issues/2#issuecomment-77372451.
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1
iQEcBAEBAgAGBQJU+IIVAAoJECjN6+QThfjzTFsIANMEUf/4y4g6e8aCcfo2bI1G 8ZKpgdC+4b1FSi1KWifJz18ZENUJ8EIEQzVQNYBPijGjyJIRgYsum63eHl271ZW1 E9/Y6Skf2OzwbQcEeEwEL4EyxmmZEZZH8qxWQWW62C+arb05Mh0+HjKJp06yh637 f4uQR6F4rR37YcQ0jU7rn8ohbWPGwSNu3NZ/SOmy++WXQoo/YGm1Qh0LTygzKPmH WaQmo7615Era0M4UrP9EohycNJxDy5zI8qMJ2K9jtTOyqUQTE4DbSeb1c4+YHNSx dOATY7yzM+cTxUINHGFzmxLJuJIiizL6Csa2UpXKvmqBAQCylRJuL9id+h6qHwI= =hd8q -----END PGP SIGNATURE-----
The idea here is to define all the disjoint classes the DBpedia ontology. This will information be used for better validation. Disjoint properties can be defined as well but the wiki does not support that yet, maybe when we move to WebProtege