On Wed, Mar 24, 2004 at 09:47:07AM +0100, Christoph Kliemt wrote: > Alun <auj@???> writes:
>
> [...]
>
> > I once worked out that my time spent managing our spam filters is
> > probably worth around ?40K per year in terms of staff time that would
> > otherwise be lost glancing over and deleting unfiltered spam. Sad to
> > say, it is economical to spend the time at a medium/large organisation.
>
> No. A false positive can be *very* expensive.
Actually humans produce more (at least an order of magnitude more!)
false positives then bayesian filters when ham/spam ratio is small.
