Re: NAF v. SNAF - where is this being addressed?

> All, forgive me if I missed something since I wasn't able to attend 
> the workshop.  My understanding from the workshop report, and from 
> discussion with Tim BL and others afterwards, was that NAF wasn't 
> going to make sense, but SNAF would -- that is, on the Web, if there 
> is not a mechanism for defining the "KB" (graph) that a set of rules 
> is applied to, there's not way to use a geenralized negation as 
> failure -- i.e. I cannot say to the "whole web" that someone can be 
> assumed to have two children unless it is shown they have a different 
> number.  Instead, I need a way to designate the dataset that a rule 
> like this is applied to.  SNAF, as I understand it, was the term 
> being used to designate this.

Jim,
Yes, SNAF is a generalization of NAF, and many people (including me) mean SNAF
(some prefer to call it scoped default negation) when they say NAF.


>   Yet, reading just about all the mail since the workshop, I haven't 
> seen this referred to at all (and it's not really discussed in the 
> WRL vs. SWRL or other threads currently being discussed in rdf-rules 
> and sws-ig)

Some systems, like FLORA-2, inherently support SNAF. WRL was supposed to
have SNAF, but not in 1.0. This is work in progress. SWSL-Rules will also
have SNAF, but not in 1.0. (These two languages are actually quite close to
each other.)

>   Seems to me if I see your rule set includes a NAF-based rule, and 
> you give me a conclusion to something, that if I don't know what 
> graph/KB/DB that was applied to, then I have no way to know whether I 
> can use your result in my application

Note that SNAF applies not only to data sets, but also to rulesets.

>    Seems to me also that this has a big effect on the charter, as I 
> don't know if there is an agreed upon use of SNAF for the Web, and 
> would need to be something the WG would be required to elucidate.

SNAF is non-controversial, I think. It is a simple extension of NAF.


	--michael  

>   -JH
> p.s. Note that in datalog, there is always the assumption that the 
> rules and a particular database can be linked - on the Web, that is 
> not necessarily true.
> 
> -- 
> Professor James Hendler			  Director
> Joint Institute for Knowledge Discovery	 	  301-405-2696
> UMIACS, Univ of Maryland			  301-314-9734 (Fax)
> College Park, MD 20742			  http://www.cs.umd.edu/users/hendler

Received on Wednesday, 29 June 2005 16:21:04 UTC