RE: Comments on "Survey of RDF/Topic Maps"

Steve

> Interested in further discussion of your point about the TMRM and the
> "(Semantic) Web."  Do you see the Web as being co-extensive with the
> range of possible subjects?  Since we all use the Web daily, it looks
> fairly flexible to us, but that is not really the same as being able
> to represent all subjects as seen by any author.

Depend on what one means by "represent". Some could think that a Semantic
Web implicit assumption is that *any* possible subject of conversation will
eventually be formally represented and uniquely identified by one (or more)
URIs providing both machine-usable subject identifiers, and hopefully
human-readable subject indicators.
IMO this is plainly impossible to achieve. Most subjects will keep being
spoken about on the Web and other information environments (publications,
data bases, media ...) without being formally represented and uniquely
identified by URIs. But various applications, using pragmatic identification
rules and algorithms, will be able to gather clusters of information, so to
speak, somehow being "about the same subject" without formal identification
or representation of this subject. Those clusters might be very transient.
Think about the top story on Google News today, under the heading "Pope
Suffered Heart Failure, Condition Very Serious".
Can one imagine that the subject of this top story is identified by the
silly URL of the all 2,497 related ...

I don't know about persistency of such dynamic URLs, and of the information
resources they retrieve. Note that there is no (human) author behind such
subjects, since Google News is driven by fully automatic search algorithms,
with no human editorial board. Would you consider Google News as a TMA,
providing it would disclose its proprietary algorithms (which I doubt)?

My point in quoting such examples is that there are much more subjects on
the Web only (not to speak about the rest of the information universe) than
the Semantic Web technologies can imagine and will ever able to formally
deal with, and that the TMRM might be a framework to deal with all those
other subjects which are not explicitly identified by URIs, but somehow
handled anyway by applications.

> As you point out, Bernard, this may not be the right venue for such a
> discussion.  It depends on the scope of the definition of "best
> practices", I guess.

That's why we should certainly port this discussion to some other venue.
Suggestions?


****************************************************************************
******

Bernard Vatant
Senior Consultant
Knowledge Engineering
bernard.vatant@mondeca.com

"Making Sense of Content" :  http://www.mondeca.com
"Everything is a Subject" :  http://universimmedia.blogspot.com

****************************************************************************
******