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

RE: Data mining the semantic web? (was RE: Semantic

• To: "'Mike Champion'" <mc@x...>, <xml-dev@l...>
• Subject: RE: Data mining the semantic web? (was RE: Semantic Web)
• From: Bill de hÓra <bill.dehora@p...>
• Date: Wed, 31 Jul 2002 09:52:35 +0100
• Importance: Normal
• In-reply-to: <XE0UTN32ZYQN0553MGLIF0ZVRPSOVU.3d4718b1@MChamp>
• Organization: Propylon
• Reply-to: <bill.dehora@p...>

> -----Original Message-----
> From: Mike Champion [mailto:mc@x...] 
>
> It would only 
> deal with these relationships and inferences *statistically* 
> whereas as I understand it RDF/OWL/etc. attempt to deal with 
> them *logically*.

Layered architectures [1].

> Or to put it another way, in this view is the semantic web an 
> ontology and trust building exercise, or a data mining exercise?

Why can't it be both? Natural language processing, knowledge
representation, machine learning, robotics, information retrieval, data
mining and rule based technologies have all prospered from looking at
hybrid techniques. The statistical versus symbolic divide was maybe
interesting to AI researchers twenty years ago, but it's a red herring
today. 

The SW effort has a penchant today for symbolic techniques (particularly
description logics); we'll see how long that lasts until someone comes
along and gets a result using probabilities in combination with the
symbolic models of TMs and RDF. 

regards,
Bill de hÓra
..
Propylon
www.propylon.com 

 

[1] http://citeseer.nj.nec.com/muller93agent.html

• References:
  • Data mining the semantic web? (was RE: Semantic Web)
    • From: Mike Champion <mc@x...>

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