W3C home > Mailing lists > Public > public-owl-wg@w3.org > March 2008

A suggestion for the resolution of ISSUE-3 (anonymous individuals)

From: Boris Motik <boris.motik@comlab.ox.ac.uk>
Date: Mon, 3 Mar 2008 13:47:29 -0000
To: <public-owl-wg@w3.org>
Message-ID: <000a01c87d35$1fb73c20$7212a8c0@wolf>

Hello,

Here is an e-mail with the proposed resolution of ISSUE-3 (anonymous individuals); this e-mail addresses my open ACTION-95.

1. Summary
----------

The idea is to extend the OWL 1.1 Structural Specification with anonymous individuals and devise a suitable semantics for them.


2. Changes to the Structural Specification
------------------------------------------

Section 4.1 would be extended with the following grammar:

<anonymousIndividual> := "the nodeID production from the Turtle document"
<anonymousOrConcreteIndividual> := <anonymousIndividual> | <individualURI>

In addition, the following text would be added:


Individuals in OWL 1.1 can be of two types: <i>named individuals</i> are known by an explicit URI, while <i>anonymous
individuals</i> are akin to blank nodes in RDF and are not given an explicit name, but are identified using a nodeID. Anonymous
individuals are local to each OWL ontology: if two different ontologies contain anonymous individuals with the same nodeIDs, these
should be treated as distinct objects, despite the fact that they have the same nodeIDs. The precise mechanism for achieving this is
implementation specific; for example, an implementation might internally prefix each nodeID with the ontology URI.


All figures containing the Individual class would be updated with the "isAnonymous" method.


2. Changes to the Semantics
---------------------------

The only thing to change is to say that N_I is the set of both named and anonymous individuals; Everything else works as expected.

3. Changes to RDF Mapping
-------------------------

In Section 2 (FS -> RDF), a row will be added that maps anonymous individuals to blank nodes.

In Section 3 (RDF -> FS), the mapping of class and property assertions will be changed to allow for blank nodes. Currently, these
are written as !x, thus requiring a resource; all these will be changed into x, thus allowing for a resource or a blank node.




That's it! Let me know what you think about this.

Regards,

	Boris
Received on Monday, 3 March 2008 13:48:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 3 March 2008 13:48:31 GMT