Re: mapping FOL

Dear Harry,
thank u,  For triple relation , i think i can understand, and in RDF
specification 1999
they have Subject->Preidcate->Object  this kind of triple relation.

well, these days, i read some paper in 2002 IEEE intelligent systems
named "DAML+OIL: An Ontology Lanuage for the Semantic Web" for some
survey study.
in that paper it mentions how to map DAML-OIL to FOL. but i am
confused when i read that part.
it shows following steps

1) translate DAML-OIL  knowledge base from its concrete syntax into
collection of RDF statement;

well, i am not really sure what is so-called "concrete syntax"

2)Translate RDF statement with Property P, subject S and Object O.
into FOL sentense of the form (PropertyValue, P S O)
where PropertyVaule denotes a ternary relation that relates a property
and an entity(a subject) to a value(an object) that they property has
for that entity.

i can not really understand this ... so maybe u can enlighten me on this.

3)Add the axioms that constrain the allowable interpretation of the
properties, classes and constraints include in RDF, RDFS adn Daml-OIL.

... so i am very clear of step 1 and 2, so .. step 3 is ... more
difficult for me to understand ...









On Mon, 27 Sep 2004 12:59:02 -0400 (EDT), Harry Halpin
<hhalpin@ibiblio.org> wrote:
> Peter,
>        Note that FOL is a higher-subset of DL (description logics), DL
> being chosen for its ability to remain decidable and having tractable
> complexity (i.e. if you send a query you get it back). However, DL is a
> bit more constrained, especially OWL/RDF DL.
> 
> In FOL book has the color red would be:
> 
> Elambda x, lambda y,(book(x), has(x,y), red(y), colour(red(y))
> which makes some assumptions about handling red and color. Not also the
> exists over lambda x and y, which meanss we have only ONE book with ONE
> color, not ALL Books.
> 
> Coding it adjectives is tricky, but I'd guess you mean that "red is a color". It
> might be different if you said "The book is bright red".
> 
> There's a million variants of FOL but that gets you the idea. And god
> knows a million ways (type/token) distinctions that get blurry, never
> mind the adjectives and subclass relations.
> 
> In DL it's less clear.
> You have to phrase it as triples:
> subject: book
> predicate: has
> object: red
> The class red is string value of the class color.
> Note you could have made it a class red is a subclass of color, but
> let's keep it simple.
> 
> So you could say (using N3 Notation)
> http://www.w3.org/2000/10/swap/Primer.html
> 
> So you want do divide things into classes and indivduals.
> Books are a class of thing, colors are a class, and red
> is a kind of color denoted by the string "red".
> 
> In the land of N3 Notation....
> 
> :book :hasColor "red"
> 
> which in full bloody glory should be qualified thus:
> 
> @prefix :<http://www.example.org/>
> @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema>
> :ColorableThing a rdfs:Class
> :book a rdfs:Class
> :hasColor rdfs:domain :ColorableThing
>          rdfs:range  rdfs:Literal
> 
> which could then be transformed into RDF/XML via:
> 
> <rdf:RDF
>  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>  xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
>  xmlns:example="www.example.org"
> 
> <rdf:Description rdf:about="example:book">
>   <example:hasColor>red</example:hasColor>
> </rdf:Description>
> 
> <rdfs:Class rdf:about="example:book" />
> <rdfs:Class rdf:about="example:ColorableThing" />
> <rdfs:Property rdf:about="example:hasColor>
>  <rdfs:domain="example:ColorableThing" />
>  <rdfs:range="rdfs:Literal />
> </rdfs:Property>
> 
> </rdf:RDF>
> 
> You could do without the colorable thing idea, but why not? You might
> want to color some things, and not color others, such as the "idea
> of loyalty".
> 
> Read:
> http://www.w3.org/TR/rdf-primer/
> 
> Cheers,
> 
>                                -Harry
> 
> 
> 
> On Mon, 27 Sep 2004, Peter wrote:
> 
> >
> > dear all,
> >
> >  given a sentence like this
> > "book has color red"
> >
> > how can i map it into  FOL format
> > subject, object, property and property value
> > which is which. thank u
> > sorry for newbie quesiton.
> >
> > yours
> > peter.
> >
> >
> 
>

Received on Tuesday, 28 September 2004 04:15:01 UTC