What would change if there were disjoint statements? Are disjoint declarations used for more than just verifying that dbpedia is consistent?
2009/7/30 Paul Houle <[email protected]>: > Perhaps the Dbpedia Ontology is restricted to OWL Lite, but I'd > really like to see some disjointWith statements in it -- I've put the > equivalent information in my system. > > One issue is that the top categories tend to be mutually exclusive: > although you can be an Actor, an Athlete and a Politician, you can't > be both a Person and a Place at the same time. I think generally the > things that are derived from owl:Thing are disjoint, I mean you can't > be a SupremeCourtOfTheUnitedStatesCase and a AnatomicalStructure at the > same time. Maybe there are cases, however, where people conflate an > Organization and a Place. > > I can think of a few places where disjointness would also be useful > below the toplevel: for instance, under Species, you can't be a Plant > and an Animal at the same time -- I think you can only be one thing at > one level in Species. Place is more complex, maybe a Mountain can be a > ProtectedArea, but a Island can't be a Cave. > > ------ > > Another general complaint is that there are things sitting around at > the toplevel that have obvious categories above them: the dbpedia > ontology does a lot better than Freebase than this (Freebase doesn't > have "Work" over "Book" and "Film"); on the other hand, it's obvious > that "OlympicResult" has "SportsResult" above it, and > "SupremeCourtOfTheUnitedStatesCase" has "CourtCase" above it. It drives > me nuts that Freebase doesn't have a base class for chemical elements > and chemical compounds -- there's no type that's good for a "Moderator" > property of a nuclear reactor since a moderator could be water > (compound) or graphite (is that an element or a compound or an allotrope > of an element or a mineral?) Seems to me that Wikipedia should have a > "Substance" category which would subsume ChemicalElement, Drug, etc... > > Is the Dbpedia ontology frozen in stone at this point or can we > still rationalize the toplevels a bit? > > ------------------------------------------------------------------------------ > Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day > trial. Simplify your report design, integration and deployment - and focus on > what you do best, core application coding. Discover what's new with > Crystal Reports now. http://p.sf.net/sfu/bobj-july > _______________________________________________ > Dbpedia-discussion mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion > ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Dbpedia-discussion mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
