W3C home > Mailing lists > Public > www-rdf-logic@w3.org > December 2000

Semantics of rdfs:subPropertyOf

From: Michel Klein <Michel.Klein@cs.vu.nl>
Date: Fri, 01 Dec 2000 17:07:07 +0100
Message-ID: <3A27CCAB.F42F5D29@cs.vu.nl>
To: www-rdf-logic@w3.org
For already a longer time, we have some problems with understanding the
exact meaning of rdfs:subPropertyOf in the RDF Schema specification. The
definition in the specification is very surprising:

    "If some property P2 is a subPropertyOf another more general
    property P1, and if a resource A has a P2 property with a value B,
    this implies that the resource A also has a P1 property with
    value B."

- First, this definition says nothing about the inheritance of the
  domain and range restrictions for a subproperty (or only very
  implicitly). Intuitively one would assume that a subproperty inherits
  domain and range restrictions from its super property. This follows
  the intuition of an is-a hierachy between classes where a member of a
  subclass is also a member of a superclass (i.e., the domain of a
  subclass is a subset of the domain of the superclass).

- Second, as [1] pointed out, inheritance of *values* creates problems in
  case of multiple inheritance. If a property has multiple subproperties
  that are attached to the same resource, it is not clear what value the
  common superproperty will have. When multiple values for the
  superproperty are allowed, we run into the problem of having multiple
  range restrictions (see the discussion [2] on this), which will make
  the reasoning on RDFS nonmonotonic.
  The above statement also seems to imply inheritance of domain
  restrictions but it is rather vague on it.

- Third, it is surprising that in the current specification values
  would inherit from sub to super properties. Again, very non-standard.

We would therefore recommend to skip the value inheritance and simply
say about rdfs:subPropertyOf:

    "If some property P2 is a subPropertyOf another more general
    property P1, this implies that the domain and range of P2 are
    subclasses of respectively the domain and range of P1."

Dieter Fensel, Jeen Broekstra, Michel Klein

[1] Wolfram Conen and Reinhold Klapsing: A Logical Interpretation of
    RDF, http://nestroy.wi-inf.uni-essen.de/rdf/logical_interpretation/
[2] http://lists.w3.org/Archives/Public/www-rdf-interest/2000Sep/0107.html
    and all reactions.
Received on Friday, 1 December 2000 12:02:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:37 GMT