Concrete types

On October 14, Dan Connolly writes:
> "Peter F. Patel-Schneider" wrote:
> >   Concrete Types (int, ...)     yes                     yes
> 
> daml-num is pretty raw.
> 
> TODO: study interaction of XML Schema datatypes (int, date, boolean,
> ...)
> with RDF and hence with DAML. see also: universe-of-discourse above.

There are a lot of issues here, and I am not sure that it is enough
simply to study XML Schema datatypes, useful though that will be. For
example, some people seem to favour the idea of defining datatypes
within the logical language, but this has many disadvantages. For
example in the case of integers, it may be difficult or impossible to
specify that there is a total ordering, and where each number appears
in the ordering. Even if this can be specified, using the
specification to infer that 1<2 may be very difficult. Moreover, there
is nothing to prevent someone asserting that 1=2 or that
(>500)=elephant.

An alternative approach (derived from DL research - see the work of
F. Baader and of C. Lutz) has been taken in OIL. Here, concrete data
types are considered to be a domain (e.g., the set of integers) plus a
set of built in predicates (e.g., <, >, =). It is assumed that the
domain is adequately structured by the built in predicates, and that
additional structuring via the concept language is not appropriate
(technically, this is imposed by assuming the concrete domain to be
disjoint from the domain of interpretation of classes and properties).

This has many advantages. For example, a standard semantics can be
assumed for integers, and there is no possibility that this can be
disturbed by strange assertions. This makes it much easier for
applications to perform inference/checking w.r.t. concrete values,
which I imagine will be a very common use of ontologies (e.g.,
products will have weights, prices etc. that should fall within
specified ranges).


Ian
--
Ian Horrocks, Department of Computer Science,
University of Manchester, Oxford Road, Manchester, M13 9PL, UK.
Tel: +44 161 275 6133  Fax: +44 161 275 6204  Email: horrocks@cs.man.ac.uk
URL: http://www.cs.man.ac.uk/~horrocks

Received on Monday, 16 October 2000 11:31:53 UTC