- From: <john.nj.davies@bt.com>
- Date: Mon, 19 Mar 2007 09:33:21 -0000
- To: <danny.ayers@gmail.com>, <public-sweo-ig@w3.org>
- Cc: <nova@radarnetworks.com>
Nova, There was a discussion recently on the SWEO list about seeing folksonomy-style approaches and semweb approaches as 2 ends of a spectrum and in no way mutually exclusive. Here's a snippet from a post I made: "I tend to present Web2.0 and OWL-based (say) ontologies as at 2 ends of a continuum, with taxonomies in the middle and thesauri, term lists, etc at different places on the line. In my view there is no mutual exclusivity here - in one recent exchange to try and get this point across I said: 'Some things (flickr) will have a folksonomy and no interest or requirement in moving to an ontology or other more formal structure. Some things (healthcare apps) will use a formal logic-based ontology and would kill people (in the case of health) if they relied on clinician's inputting semi-random tags. Some things (some semantic wiki apps, for example) will benefit from marrying Web 2.0 and semantic technology.' " And here's some very good (IMHO) slides from Jeff Pollock @ Oracle: http://www.w3.org/2001/sw/sweo/public/2007/Pollock.SDForum.Ontology-vs-F olksonomy.small.ppt Regarding non-trivial use cases of semantic technology, there is a 400000+ class OWL ontology representing the SNOMED-CT medical terminology (http://www.snomed.org/) which we in BT are using with customers right now if that helps. As I mentioned baove, if you tried to use a folksonomy for the sme thing you might end up killing people. Regards, John Davies. -----Original Message----- From: public-sweo-ig-request@w3.org [mailto:public-sweo-ig-request@w3.org] On Behalf Of Danny Ayers Sent: 18 March 2007 18:22 To: W3C SWEO IG Cc: nova@radarnetworks.com Subject: Fwd: Question... [seems like a job for SWEO!] ---------- Forwarded message ---------- From: Nova Spivack <nova@radarnetworks.com> Hey guys - I'm supposed to be speaking about the semweb on a panel moderated by none other than Clay Shirky (who hates the semweb and generally misrepresents it) on Monday (at the Highlands Forum). I was wondering --- are there any stats I can cite about the semweb that will help to head of his criticisms? - Number of SPARQL endpoints, or growth of them? - Growth of semweb? Any metrics you know of? Anything else you think I might cite? Clay generally attacks the semweb on the grounds that social tagging is better than inferencing. Duh. That's a completely made-up conflict since the semweb is a natural fit with tagging, and secondly since the semweb does not only focused on inferencing. In any case, the question is how to deal with him. N -- http://dannyayers.com
Received on Monday, 19 March 2007 09:33:29 UTC