[Z3-zemantic] Re: Zemantic 0.5 released

[Michel Pelletier]

On Tue, 2005-07-19 at 12:29 +0200, Olivier Grisel wrote:
> Michel Pelletier wrote:
> 
> > Nothing on this particular subject yet.  More recently I've been working
> > on RDFS entailment in rdflib which is a priority right now,
> 
> Interesting. What kind of 'entailement' are you working on ? RDF 
> validation against a schema?

RDF and RDFS entailment as described in the W3C document 'RDF Semantics'
at:

http://www.w3.org/TR/rdf-mt/

It's a very technical document, but in particular see section 7.
Basically there exist through proof a set of rules and standard RDFS
axiomatic triples that are a consequence of RDF and RDFS's model.  For
example, if you know that X is a rdfs:subProperty of Y, then you can
'entail' that both X and Y's rdf:type is rdfs:Property.  Much of the
theory beyond that is beyond me. ;)

Many of the rules can be applied recursively and some even infinitely so
it's a tricky issue to solve.  I'm working on a general framework in
rdflib that will allow users to plug in entailment policies that are
specific to their application.  There will be provided a simple "poor
man's" policy that does the basics like property and class entailment.

Libraries like this will assist agents and search languages to derive
new information from a graph that may allow them to solve a particular
goal or find certain query results.


> Yes that what I thought. But the use of some semantic base (a RDF model 
> of some kind of english dictionary) would allow us to use to do semantic 
> inference on the content of the indexed document with a knowledge of 
> synonyms, antonyms, ... This would require to some kind of structured 
> thesaurus. The OpenOffice project has some GPL thesaurus that may help 
> to build this. Wordnet is probably more advanced, but not free ... oh 
> wait, it looks like it's BSD-like now: http://wordnet.princeton.edu/license
> So Wordnet is probably the way to go.

I think this is a great place to start and would love to help out on
that.  BTW if it's important I'd like to point out that NLTK is GPL.

> We would also probably need to define some kind of semantic distance 
> between the predicates of our common sense ontology and then be able to 
> do appromatively correct queries on the model using a threshold on that 
> distance. I'm not really aware of the research results on this kind of 
> problems. Should you have any pointer on the topic, please feel free to 
> share them :)

You lost me on this one.  If there is an existing ontology then we
should just skip devising the common sense ontology.

> Yes, I guess we first need to work on some toy examples to work out what 
> works and what's pure utopia. I will first have a look at the NLTK 
> tutorials to get a better picture of what it actually does. BTW, do you 
> think we should use NLTK-lite that looks more pythonic than the 
> tradional NLTK?

Good ideas all, I've never heard of NLTK-lite, but NLTK is written in
pure python, no?  What could be less pythonic than that? ;)

-Michel