- From: Charles McCathieNevile <charles@w3.org>
- Date: Tue, 18 Sep 2001 05:07:15 -0400 (EDT)
- To: Al Gilman <asgilman@iamdigex.net>
- cc: David Poehlman <poehlman1@home.com>, wai-ig list <w3c-wai-ig@w3.org>
Charles, Danbri, Libby, Marja, and the "Dublin Core community" among others. In discussing SVG as a leap forward for accessibility, a major part of it is the ability to pull apart and put together known pieces of imagery http://www.w3.org/TR/SVG-access (It even allows this to be done for formats thought of as less useful, like .gif and .jpg) In looking at the benefits of the semantic web, we see the other piece of the puzzle - the ability to record these linking statements http://www.w3.org/2001/Talks/0627-semweba has some little examples The annotea project allows us to make statements about anything on the Web, including RDF explicitly relating an image to a word (for example derived from wordnet, a dictionary that itself makes formalrelationships between words, such as "the noun 'eagle' refers to a subclass of 'raptor' - and that is self is a type of 'bird'"). Dublin Core is a small vocabulary of statements about things, and a vast amount of information is collected in dublin core statements. One of the many tools for doing this is photo-rdf http://www.w3.org/TR/photo-rdf which adds information directly into jpeg files (which is the typical format for photographs). Dan and Libby have been working on a tool for recording "codepictions" - pictures of two or more people ( http://www.w3.org/2001/08/rdfweb/svg-foaf ). I added to that a technique for marking which bit of a whole image is a picture of which person - easy really, because it simply involved using the ideas written about SVG and semantic web coming together. These are a couple of home-grown solutions - rather like http in that sense. If nobody uses them, they die, if people pick them up they have potential for providing masses of information. There are of course other such systems, and the trick is to link them. ==== technical philosophy discussion: Why RDF in particular and the semantic web in general? Well, RDF in particular is a format for recording semantic web information where there are lots of available tools that can be re-used. It is like the choice of HTML for web pages - it has a lot of working tools, and some good features. It probably isn't perfect yet, any more than the alternatives, but it seems likely that it has critical mass to become a standard (as in "the thing that everyone uses", not just "the thing we said people should use"). The semantic web is a way of collecting up the various bits of information without having to have a single massive server, which would require a lot of resources. It has the downside that collecting up information that is initially very scattered is slow. But as the connections are made, the rate of information available grows extremely rapidly - connect two services together of the same sizes and you double what is there. And so on... (well, this is how the Web grew to a fairly large extent). On Sat, 15 Sep 2001, Al Gilman wrote: OK, let me give this away. This is worth an NSF grant, in my opinion. I had been dog in the mager hoarding it on that basis. [snip] The basic resource is a thesaurus which relates words and pictures. What [snip] There is is. it takes mobilization of an organized team to pull it off and maintain the hub resources. Who wants to make it happen?
Received on Tuesday, 18 September 2001 05:07:19 UTC