Jeni's Musings

Finally got around to blogging this -

http://danbri.org/words/2009/01/18/390 Re claims/evidence design, see http://esw.w3.org/topic/HCLSIG/SWANSIOC for efforts on modelling scientific discourse.

And w.r.t. modelling ‘personas’ (occurances of people), there’s a series of posts around http://www.r4isstatic.com/?p=26#comments that describe efforts at the bbc to describe fictional people and works of fiction. While this might seem another problem (and it is!) there are some common themes - for example picking out occurances of the same real-world person even when they occur in a work of fiction (eg. Mark Twain in Star Wars). This needs some way of saying “these layers of data don’t directly describe the world, they describe the world-according-to-some-text” (or “content”, or whatever). But doing so in a way where the entity identifiers can still be cross-referenced with other sources.