- From: Dr. Francis MUGUET <muguet@ensta.fr>
- Date: Fri, 07 Oct 2005 23:37:04 +0200
- To: rich@boakes.org
- CC: Danny Ayers <danny.ayers@gmail.com>, semantic-web at W3C <semantic-web@w3c.org>
Dear Rich > There is no such thing as clean unspoiled metadata. Here we are ! a case of semantic misunderstanding ! I meant spoilt, junk, corrupted Metadata, not unbiased metadata, in fact raw text is also biased... We plan to use automatic junk busters to detect sites where data are so corrupted that they jeopadize the whole system as it is the case now in the World Wild Web. search engines are pretty good harvesting biased text files, and their expertises would be quite interesting for SWgTLDs. > > Anyone who writes metadata, anyone who annotates > content, and in fact, anyone who creates the > content in the first place, has a reason for > doing so: *everyone* is biased. agreed ! > > The important thing with metadata is it's > provenance: Your point gives support to my proposal. With a domain name, we can know, the provenance ie the owner of the domain, to a better extent than with a certification authority. If the owner appears to be a dummy or a non-existent person, we kick the domain off the registry. Again, we are not a commercial venture, we are not looking for customers, building trust is more important. Then being a site with a SWgTLD is going to be the mark of an overall provenance. However, provenance is only one ingredient, search engines experience would be quite interesting to use within a SWgTLD. if you know the origin of a piece of > information, you can decide how much you wish it > to affect the representation of the web that you > see. A domain name would a step forwards in that direction. > > All that spoiled, biased, metadata is what makes > the semantic web different from the web. Where > once there were links, now there are machine > interpretable opinion. and junk. The problem is that junk is even worse for machines, when they do not know what it is junk... ( ontology of the junk... ) Best regards Francis -- ------------------------------------------------------ Francis F. MUGUET KNIS/ENSTA Pôle de Développement pour l'Information Scientifique "Réseaux de la Connaissance et Société de l'Information" Scientific Information Development Laboratory "Knowledge Networks & Information Society" (KNIS) ENSTA 32 Blvd Victor 75739 PARIS cedex FRANCE Phone: (33)1 45 52 60 19 Fax: (33)1 45 52 52 82 muguet@ensta.fr http://www.ensta.fr/~muguet mirror http://www.muguet.org MDPI Foundation Open Access Journals Associate Publisher http://www.mdpi.org http://www.mdpi.net muguet@mdpi.org muguet@mdpi.net World Summit On the Information Society (WSIS) Civil Society Working Groups Scientific Information : http://www.wsis-si.org chair Patents & Copyrights : http://www.wsis-pct.org co-chair Financing Mechanismns : http://www.wsis-finance.org web UNMSP project : http://www.unmsp.org WTIS initiative: http://www.wtis.org ------------------------------------------------------
Received on Friday, 7 October 2005 21:27:20 UTC