Re: RIF: A thought about requirements --> PRR

From: "Gerd Wagner" <wagnerg@tu-cottbus.de>
Subject: RE: RIF: A thought about requirements --> PRR
Date: Fri, 2 Jun 2006 09:33:06 +0200

>>a RIF that covers logical as well as production rules would:
>> - help SemWeb technologies enter the market mainstream 
>> - and it would help mainstream (in a market sense) rule 
>> technologies enter the Web, including the Semantic Web 

> I believe that this depends on *how* the RIF would cover both 
> logical and production rules. If the RIF is in essence multiple 
> formalisms sharing at most a portion of the syntax (with the 
> production rule formalisms being disjoint from the formalisms 
> that cover logical rules and the existing W3C Semantic Web 
> languages), then I don't see how it would  advance either of
> the above goals.  

> Why can't you see this?

Well, I don't see this because I don't see how an interchange language
that doesn't interchange between logical and production rules is going
help push either logical or production rules into the sphere of the
other.  Where is the mechanism that would provide any force for either
of these movements?

> Because you don't want to see it
> (and just want to push your Unitary Church of the Holy Owl)?

I do not believe that this is the case.

> The precedence is SQL, which has several sublanguages
> catching different formalisms: SQL views correspond to
> constructive derivation rules, SQL assertions correspond
> to normative/integrity rules, and SQL triggers correspond
> to ECA/reaction rules. These sublanguages share some syntax
> and semantics (which is quite useful), but represent
> different formalisms.

Yes, SQL includes various different kinds of things in an overall
framework that has some utility.  However, the RIF and SQL are
different kinds of things.  SQL is an actual language and is seen and
used by quite a number of people.  The sheer intellectual effort of
having to do anything in SQL leads to a certain understanding of the
differences and commonalities between the various parts of SQL, at
least in some people.  This can lead to some beneficial forces.

However, the RIF is not anything that will be used directly and thus
will not provide this kind of force.

If the RIF did indeed require significant commonality between the
different formalisms that used it then the RIF would indeed provide
some forces to aid in either of the two goals above, but in the
absence of such commonality (and, no, I don't consider simple partial
syntactic commonality to be significant commonality) I just don't see
any resultant forces towards helping production rules enter the
Semantic Web or helping SW technologies enter the market mainstream.

I do see that syntactic commonalities, i.e., a web-based syntax that
covers production rules, may slightly help production rules enter the
Web, but beyond that I just don't see.


peter

Received on Friday, 2 June 2006 09:39:08 UTC