[FoRK] Science Braces for Second Term

Eugen Leitl eugen at leitl.org
Wed Nov 17 12:21:15 PST 2004


On Thu, Nov 18, 2004 at 03:43:46AM +0800, James Tauber wrote:

> One, for example, might trust Wikipedia more because of its wiki-nature (and
> the kind of self-regulation that results) than the credibility of any
> individual contributor.

Quality of Wikipedia entries will vary over time and topics covered, as well
as depend on current process governing how information is retained. Of course an article
covering a region in knowledge space is better than this region to remain
immacolate of ink. But, as soon as we have a lot of churn in an article, trust
becomes an issue. Gross defacement and vandalism is easily detected and corrected, assuming
there's a tracking and rollback mechanism. What about more clever form of sabotage, though? Or "improvements"
done in good faith but which degrade the article?

Trust is relative, so you need to store the article editing history as a time
trajectory, including branching, and transparently navigate here using trust metric
clustering, which is, thankfully, comparatively easy (unless someone fakes
warm bodies on the other end, and foils the clustering function with
judicious insertion of garbage).

Most of this is academic for obvious reasons, but measuring and asserting
quality for different audience targets, especially in a hostile publishing environment,
is a nontrivial/unsolved problem.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net


More information about the FoRK mailing list