Comment on RDF comment

I believe that the lack of "real" comments in RDF is a crucial
omission that makes it unsuitable for use in realistic ontological
engineering, or as as the basis for other languages (such as OWL) that
will be so used.

The point is this. RDF/OWL are supposed to be ontology languages, and
users will (hopefully) want to use them to build ontologies. In many
cases these will be large and complex, and will be worked on by many
authors using many different tools. They will hopefully also be
shared, extended etc. As in software engineering, the use of comments
will be important in this ontology engineering process.

Such comments are not intended to have any impact when the ontology is
used in applications (in the same way that comments have no impact on
code when it is compiled and run). In particular, it would be
inappropriate for applications to infer semantic differences in
information represented in two ontologies based solely on differences
in comments (in the same way that it would be inappropriate for code to
behave differently when only the comments are changed).

It has been suggested within the WebOnt working group that the use of
XML comments is the solution. This does not work, however, as XML
comments would be stripped by RDF parsers, and so comments would be
lost whenever an ontology goes through a read/write cycle, e.g., when
being edited (if I can be forgiven for banging away at my software
engineering metaphor, this would be equivalent to emacs stripping
comments when editing source code). Moreover, there is no standard
mechanism for indication which component(s) of the ontology an XML
comment refers to.

As comments are a "must have", implementors of ontology tools will be
forced to adopt ad hoc solutions. This will severely impede tool
interoperability. Tool interoperability is one of the main benefits
that will encourage the initial uptake of language standards such as
RDF and OWL, and lack of interoperability may cause potential users to
look for solutions elsewhere.

On the other hand, I can't see how making rdf:comments be "real"
comments, i.e., have no semantic impact, would have any adverse effect
on RDF.

Regards, 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 6211  Email: horrocks@cs.man.ac.uk
URL: http://www.cs.man.ac.uk/~horrocks

Received on Thursday, 20 February 2003 18:29:24 UTC