update to MT document

An updated version of the MT document can be found at

http://www.coginst.uwf.edu/users/phayes/w3-rdf-mt-draft-J.html

This is not *quite* finished, in that two of the proofs are not yet 
completely written out (marked by @@@@). Also I  may need to re-word 
the rdfs closure lemma slightly. But as I may not have time today to 
get this done, and since most of the changes can be checked 
independently of the math fine-structure, I thought I would pass it 
around in this state for eyes-overing. I should get the proofs done 
tonight and a publication version ready and code-checked by Wednesday.

The main changes are styled in yellow (to remove, change the 'new' 
style in the header). Ive made several substantial changes to the 
document, briefly explained here.

1. Added some explanatory purple prose going on about the nature of 
interpretations, what entailment really means, etc., and made several 
prose changes, mostly following Graham's suggestions, eg 'identical 
RDF graphs' instead of 'the same RDF graph'.. (Also, got rid of (most 
of) the brackets).
2. Trashed the completely unreadable Appendix on graph definitions, 
and re-written the account in the main text. I think this is good 
enough for any reasonable reader, and the long stuff was just a 
distraction. After a month, I couldn't follow it myself.
[In this version, graphs are not tidy on literals, which is the most 
'conservative' assumption for the MT. That can easily be changed, if 
required, in later versions, with small changes to the rest of the 
document. ]
3. Reorganized many of the graph definitions and put them all 
together into a new section 0.3
4. Put links back from technical terms to their definitions in many 
parts of the document, as suggested by Jeremy, particularly where the 
first use is remote from the definition. [Hypertext lives!]
5. Made a few technical changes, mostly arising from actually proving 
the lemmas, some of which have had to be re-worded slightly to avoid 
the counterexamples noted by Jeremy and Graham, which in turn 
involved having some extra definitions.  Ive stated a few extra 
lemmas, but I know they are in fact useful because I needed them for 
the proofs of the other lemmas; and Ive re-ordered the material in 
that section to make it flow better.
6. One slightly important technical change is the removal of one of 
the conditions on an rdfs-interpretation, concerning rdfs:Literal. 
Without this, the rdfs-entailment lemma would fail, it turns out, 
since there is no way to write a closure rule that forces things with 
type 'rdfs:Literal' to actually be the values of literals. I've put a 
comment about this in section 6.1. This is the only substantial 
change to the MT .
7. Deleted section 7 on containers until we get that sorted out.

Pat

PS. I havn't done all the changes suggested by the readers. The one I 
feel bad about is not altering example 1  to be more 'realistic' as 
requested by Patrick, and which I promised to do. That is a 
combination of laziness and having lost my copy of the 
graphical-drawing software. If I promise to do that for the final 
version, will this be OK for now? Basically, I want to get the 
correct RDFS rules, and the idea of vocabulary entailment, out into 
the public arena as quickly as possible.
-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Monday, 28 January 2002 13:23:11 UTC