Re: What do the ontologists want?

>
>
>Seconded: We are in a similar situation with agent-based command
>information systems; having an audit trail and being able to timestamp
>and attribute statements is absolutely essential here, especially as we
>are likely to encounter deliberate misinformation which must be filtered
>out at a later time.
>
>For a local database we could achieve this 'outside of RDF' using a
>custom framework, but we might want to be able to pass this attribution
>information to and from other parties, in a standard way. Is there a
>convenient way of doing that without reification?
>
>Regards,
>
>David Allsopp
>DERA Malvern
>UK


actually, although we've been discussing this a lot in DAML+OIL 
committee (under the name tagging) I've come to realize that there's 
a better way to do this in the ontology world -- define an ontology 
in which objects have these properties and use them -- the languge is 
then in the standard, but different communities can work on different 
tagging schemes

Essentially, create a DAML ontology with

Agent:Message :a daml:class .

Agent:TimeStamp :a daml:property
   rdfs:range  Agent:Message
   rdfs:domain <define your own or use XML datatype>  .

etc.


Later, you use instances of these in the obvious way

Agent:Message11111988 :a Agent:Message
  Agent: TimeStamp "4:00PM Tuesday"    [or your favorite]  .



You can now have relatively arbitrary data structures, they're in a 
relatively stable language parsable with either RDF or XML tools, and 
you can exchange things freely with colleagues using various DAML or 
RDF tools.



I'm working on one of these for exchange of proofs, and there are 
other ways to go.

-- 
Prof. James Hendler		Program Manager
DARPA/ISO			703-696-2238 (phone)
3701 N. Fairfax Dr.		703-696-2201 (Fax)
Arlington, VA 22203		jhendler@darpa.mil

Received on Tuesday, 15 May 2001 18:59:32 UTC