[net.news] Keywords - tolerance for false hits

hes@ecsvax.UUCP (Henry Schaffer) (10/21/85)

> > ... my earlier posting on foreign language technical material 
> 
with distribution net.ai,net.arch,net.math,net.graphics,net.physics,net.bio
(and maybe more)
> 
> What has this garbage got to do with mathematics? Are you sure that 11
> different newgroups are enough to post the same message to? Maybe you'd
> like to try for 20 more??  Please restrict postings to this group to
> mathematics and closely related areas.
> 
  This is an example of how strongly people can feel about receiving an
article which they don't feel belongs to a group.  Would people feeling
like this be satisfied with a keyword system which gives something 
like 20-30% false retrievals?
  I've never seen a retrieval system which has as high a percentage of
relevant material as net.math has of mathematical articles, and further
one might even consider that the first article (on translation of 
foreign language technical material) to be relevant to mathematics.
--henry schaffer  n c state univ 
Note: I've removed the names of the posters, since I'm not discussing them.
I'm trying to make a general point.