xml-dev - Re: Semantic Web Hackings

["Sean B. Palmer" <sean@mysterylights.com>]

> Danbri did some stuff a few months ago

Yes, in his FOAF Schema, there is the following property:-

     http://xmlns.com/foaf/0.1/mbox
     A web-identifiable Internet mailbox associated with
     exactly one owner.

I'm not sure how you'd tie that into the single person <=> single URI
dichotomy that I've been talking about though, becasue it would be using
the mailbox URI itself as the property value... maybe Dan could invent a
new property to cope; what we need is a "mboxOf" property, that asserts
that a mailbox has exactly only one owner, whos name is "x". In fact, all
properties should have an inverse property, and I suppose that "mboxOf" is
the inverse of "mbox". Example usage:-

     <rdf:Description rdf:about="mailto:sean@mysterylights.com"
        foaf:mboxOf="Sean B. Palmer" />

> There is stuff in DC that could probably be sensibly applied

"dc:description" springs to mind:-

     <rdf:Description rdf:about="mailto:sean@mysterylights.com"
        foaf:mboxOf="Sean B. Palmer"
        dc:description="This is the Web mailbox object of a single person
          who's name is x"./>

> I am not sure that there is an intrinsic need to disambiguate
> people perfectly

Well, in some cases there is: I don't want anyone else using my credit
card; but somethimes there isn't: one mailbox may be owned by a company.
Even then, there is still the fact that "someone" must own a Web box. If
it's owner can be verified and is unique, then why not use it as a
representation of the digital interface of that person? Hence you say:-

> it is more a case of being able to establish how much you trust
> a particular source

Precisely. In this case, the source is a mailbox of a person / group of
people / company etc. Becuse a URI is unique, and a name isn't as long as
you have 1) A mailbox 2) a person 3) a digital signature, you can say that
you want to transact with 1) as long as 3) is verifiable, and then give the
whole system the alias of 2). This is very strikingly close to some of the
stuff that TimBL talked about when describing the Semantic Web in XML2000.

> I am pretty sure that danbri's stuff uses mailto: and web pages as
> "tags" for a person. If you add a date to a statement then you can
> say "{mailto:charlesn@sunrise.srl.rmit.edu.au in december 1997}
> refers to {mailto:charles@w3.org in December 2000}" and then
> that either of those is the same as ...

Yes: this is the Ontology and Logical layer that I was talking about. There
are quite a few things you could do with this mailbox/FOAF stuff:-

1. If p1 has name A and p2 has name B, then p1 != p2
2. If p1 has age of 25, then p1 does not have any other age (until next
year)
3. p1 has mailbox xxx1, which used to be mailbox xxx0
4. p1 can be trusted
5. If p1 can be trusted, and p1 trusts p2, then trust p2

And so on. Endless possibilities! The best thing is that this can all be
done *now*. We don't have to wait for a Semantic Web to come along before
all of this will work. I could use Dan's FOAF stuff, and the uniquness of
my IP to assert that:-

1) mailto:sean@mysterylights.com represents Sean B. Palmer
2) This is a definitve address as of todays date
3) This URI (that represents me) has these properties
4) You may make logical assumptions about these properties if that is
allowed by their Schema.

Problems: Do I have to backlink if I no longer control that server/email
box? All of these current properties are reified...is there no way to make
some more sense of all of these using other RDF principles? What is the
most secure way to assert that a URI is under the control of person x? What
practical uses are there *today* of making SW assertions about a
mailbox/person?

Kindest Regards,
Sean B. Palmer
http://infomesh.net/sbp/
http://www.w3.org/WAI/ [ERT/GL/PF]
"Perhaps, but let's not get bogged down in semantics."
   - Homer J. Simpson, BABF07.