On Sat, 23 Mar 2002, Aimee Farr wrote: > The real issue seems more properly couched as salience. The blur here causes > conceptual errors, and I would appreciate enlightenment, by way of an > alternative taxonomy and any refs to recent papers measuring the "S/N ratio" > within a channel.
There's no observer-invariant ranking of content. There's clustering, though. User feedback paraphernalia attached to each message (say, X-Rank: http://cpunx.org/cgi-bin/rank?agent=farr&rating=doublepluspunkalicious), similiarly to Googlebar's smiley/frowny and a bit of clustering run server-side could do wonders here. Trouble is, you'd need MUAs who are aware of X-Rank, and/or attach ranking urls at the bottom of each message. Plus, server-side infrastructure.