[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

metadata from any source (was RE: RE: Semantic Web)

• To: XML-DEV <xml-dev@l...>
• Subject: metadata from any source (was RE: RE: Semantic Web)
• From: Edd Dumbill <edd@u...>
• Date: 31 Jul 2002 14:30:27 +0100
• In-reply-to: <8BD7226E07DDFF49AF5EF4030ACE0B7E0696F6CE@r...>
• References: <8BD7226E07DDFF49AF5EF4030ACE0B7E0696F6CE@r...>

On Tue, 2002-07-30 at 22:52, Dare Obasanjo wrote:
> This is paraphrased [and mangled] from conversations with Joshua Allen
> who brings this up every time I talk to him about the SemWeb; "A
> semantic web where only metadata published by the owner of the content
> is available is next to useless. The value of the SemWeb is being able
> to obtain metadata about content from any source" 

I've had some small success implementing this with FOAFbot, see
http://usefulinc.com/foaf/foafbot/

This is an IRC bot that spiders RDF descriptions in the form of FOAF
(see http://rdfweb.org/foaf/) files and attributes them to an owner
based on a PGP signature.

Here's a simple sample session:

<edd> foafbot, edd's email
<foafbot> edd's email is '[mailto:edd@u...]', according to Dave
Beckett, Edd Dumbill, Kip Hampton, Matt Biddulph; and anonymous sources
Anon12, Anon2, Anon37, Anon41
<foafbot> edd's email is '[mailto:edd@x...]', according to Edd
Dumbill; and anonymous sources Anon19

The current downside is that I'm implementing provenance tracking on top
of my chosen RDF toolkit (Dave Beckett's Redland), which makes things
rather slow.  Work's underway to bring such tracking down to the toolkit
level.

-- Edd

This is a digitally signed message part

• References:
  • RE: RE: Semantic Web
    • From: "Dare Obasanjo" <dareo@m...>

• Prev by Date: RE: The BBQ Proposition was Re: Re: URIs, concrete
• Next by Date: RE: Data mining the semantic web? (was RE: Semantic Web)
• Previous by thread: RE: RE: Semantic Web
• Next by thread: RE: RE: Semantic Web
• Index(es):
  • Date
  • Thread