libby.miller at b...
Tue Mar 5 16:55:57 UTC 2002
On Tue, 5 Mar 2002, Charles McCathieNevile wrote:
> Well, I don't know if "number of people in a photo" is a good metric - among
> other things it is not clear that the people listed as being in a given photo
> are all the people in that photo...
very true. It might work as something approximate though.
> but it might be intersting to look at different weighting mechanisms for the
> strength of a link.
danbri and shellac and I were just discussing various ways of doing
this, and various metrics. amihotornot springs to mind; classifying for
nudity (or lack of) as well ;)
> On Tue, 5 Mar 2002, Libby Miller wrote:
> cc'ing damian, who did most of the coding!
> hi max
> interesting idea, and of course the number of (described at any rate)
> people on the photos is somthing we already know. we'll have to have a
> little think about implementation - damian is probably the best person
> to do that :)
> several people have been interested in differing shortest paths, but
> noone's come up with this idea, but I think you're right that a smaller
> number of people is better 'quality' somehow, maybe because they're more
> likely to know each other better.
> I wonder if we should direct this to the rdfweb mailing list? (cc'd, I
> On Tue, 5 Mar 2002, Max Froumentin wrote:
> > While zoning out during the delivery context workshop I had a idea
> > for making codepiction. Someone may have had this idea before, in which
> > case just ignore this.
> > The idea is to introduce link "strenghts" to photos. A photo showing
> > me with chaals among 50 people has less value than one where there's
> > just the two of us. Indeed I'm more interested by the following path:
> > maxf -> danbri -> chaals (where there's few people on each photos),
> > rather than maxf -> chaals on a W3C Team group photo, even if the
> > former path is shorter.
> > In terms of implementation, it's just a matter of assigning a weight
> > (or distance) to links from the number of people on the photo, and
> > compute the shortest path (using Dijksta's algorithm).
> > Max.
> Charles McCathieNevile http://www.w3.org/People/Charles phone: +61 409 134 136
> W3C Web Accessibility Initiative http://www.w3.org/WAI fax: +1 617 258 5999
> Location: 21 Mitchell street FOOTSCRAY Vic 3011, Australia
> (or W3C INRIA, Route des Lucioles, BP 93, 06902 Sophia Antipolis Cedex, France)
More information about the foaf-dev