nodalpoint.org - Rethinking the Semantic Web part 1 - Comments

Neil — Tue, 20 Dec 2005 23:23:41 -0500

I must admit, discussions about the semantic web leave me baffled. I get the idea - that we use metadata to deliver relevant content rather than relying on keyword searches. However, reading articles like those mentioned leaves me with no sense of how this is being achieved (if it is). It seems to be an almost deliberately abstruse and jargon-riddled field - jargon for jargon's sake - which, for people whose goal is information delivery seems rather absurd.

I'd like to see some practical demonstrations, proof of concept, particularly with relevance to biology. Perhaps such sites exist? Set up a server with some content, write a prototype client and show me how it works and the advantages over what I already use. Otherwise it's all just so much highbrow intellectual arm-waving.

Rather interesting piece. It

Greg — Tue, 20 Dec 2005 07:05:26 -0500

Rather interesting piece. It is nice that Rob McCooool has offered up his reasons why the semantic web won't work, I mean *who hasn't* criticized the semantic web at some point ? It is becoming like a club or something, "The semantic web is a pipe dream and won't work because of...". Funny thing is I don't think TBL at any point has said the semantic web will be an easy thing to achieve... of course Rob has a solution, doesn't everyone have a solution to the semantic web ? but I think the contribution is not so much his solution but *how* the solution is presented.

Think about it, the article is entitled "Rethinking the Semantic Web part 1" and the solution is presented in the "To be continued..." part II. Pure genius, you just can't help yourself begging for more ? So from this point on I plan to publish all papers in in either two parts or three. Part one will will have the introduction, materials and methods and a results teaser... just to keep you on the edge of your seat for part II, full results and discussion. I haven't figured out the trilogy format yet.

Seriously though.

I am curious about his proposed solution as I agree with many of the failings he states e.g. how OWL deails with exceptions and context. This is close to my current issue with the semantic web, that people miss the point that It Takes an Agent to be Semantic. It is not just about controlled vocabularies and reasoners.

However for someone claims to have a solution to the semantic web's problems there are some odd statements in the article. For example "RDF triples must be maintained, either as separate files or within separate blocks inside HTML files." What about in a triples database ?

"The ontological data model makes representation of any nontrivial factual information difficult because it can’t represent context of any kind." What about named graphs ? What about the fact that context (in the semantic sense and not just the provenance of the triple) is dependent on shared meaning (common knowledge) ?

See Kashori (John Black) for more on how the semantic web is likely to emerge.

Rethinking the Semantic Web part 1

Duncan — Fri, 16 Dec 2005 12:49:44 -0500

As it stands, the Semantic Web will not achieve widespread adoption, according to Rob McCool of Yahoo! / Stanford in his article Rethinking the Semantic Web, part 1 published in the nov/dec issue of IEEE Internet Computing. Rob outlines several reasons for this, but we have to wait for part 2 to see Yahoo's solution, which will be published in January 2006.

nodalpoint.org - Rethinking the Semantic Web part 1 - Comments

See also: The problem with

proof of concept

Rather interesting piece. It

Rethinking the Semantic Web part 1