class-specific properties considered harmful

[VladimirAlexiev]

Consider these properties:

• http://dbpedia.org/ontology/Galaxy/apoapsis
• http://dbpedia.org/ontology/Planet/apoapsis
• http://dbpedia.org/ontology/Spacecraft/apoapsis

They mean exactly the same and the unit is the same (km). They just pollute the namespace (eg if you look in http://ismaro3.ddns.net/webprotege, you'll see a bunch of props with the same name and wonder which one to use) and add unnecessary complexity.

apoapsis_km could make sense. But Galaxy/apoapsis makes no sense at all.

Class-specific properties are supposed to provide more natural units, eg "cm" for person height as opposed to "m" for height, which is the default. But (observed in http://vladimiralexiev.github.io/pres/20150209-dbpedia/dbpedia-problems-long.html#sec-7-5):

• you need to know such per-class property exists
• you need to know the unit (eg see http://dbpedia.org/ontology/Person/height), whereas the universal props use SI
• you cannot use them with prefixes because eg dbo:Person/height is not a valid prefixed name. This should be changed to eg dbo:Person_height (or even better dbo:height_Person)

Because of the inflated complexity and the inconveniences described above, I propose to remove these props, and replace them with unit-specific props, eg apoapsis_km, height_cm

[reeshabhranjan]

Is this still open and worth to be worked on?

[JJ-Author]

Since the feature seems rarely used it does not really make sense to work on it or the impact would be really low. One contribution could be to actually analyze the usage in order to quantify it.