[topicmapmail] Regarding Facets

[Murray Altheim]

Thomas B. Passin wrote:
> [Nagarjuna G.]
> 
>>I wanted to understand the reasons why facets are not included in the
>>XTM standard?  If we eneed to add some special properties of topics (say
>>a book has an accession number) how do we represent in XTM?
> 
> In XTM there are occurrences, which belong to specific topics.  If you index
> the occurrences by their types, the index is essentially a collection of
> facets, I think.  An accession number would be put into an occurrence of the
> topic that represents the book.

While an interesting approach, this is really overloading the idea
of topic occurrences, since in your suggestion the facets aren't
really occurrences of the given topic but facets (properties) of
it. But since this kind of hack is simple and likely a proprietary
usage, I don't doubt for a minute people haven't or wouldn't do it.

As Steve Pepper added, the more semantically correct way is to do
this is to treat the information resource as a topic (not using the
original topic) and associating it to the original as a facet, i.e,.
asserting it as an instance of a facet.

Nagarjuna added in a followup message:
 >Another associated problem: If we want to contrain a valid data type
 > (number, string, etc) for the values of an occurance, how do we do
 > this.

This problem hasn't actually been solved for topic maps yet, though
I've published a draft solution at

   http://kmi.open.ac.uk/psi/datatypes.html

which essentially steals the XML Schema datatypes semantics so that
you could apply them in a topic map. At some point I hope to get
feedback on this document and edit both the spec and the associated
topic map to conform to the PubSubj TC's recommendations on
publishing PSI sets.

Topic Map applications that supported these datatypes would allow
constraint checking to be performed. This type of validation would
be considered a useful but domain-specific solution, not part of
topic maps themselves, sort of like a first order logic PSI set
would allow logical expressions within topic maps, but you'd
still need an inferencing engine to do the actual validation --
there's no magic here.

Murray

......................................................................
Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK

     "In Las Vegas Mr Gates also demonstrated a prototype
      fridge magnet which can be programmed to receive traffic
      reports, sports results and advertisements from local
      restaurants using the same FM signal as the wristwatch."
                                  -- The Guardian, 10 Jan 2003.