On 29/12/2008, Brian <Brian.Mingus(a)colorado.edu>
wrote:
> So why are you wasting the ISPs time and the police's time when the best of
> the passive technology routes have not been explored? Using machine learning
> *you pit the vandals against themselves. *Every time they perform a
> particular kind of vandalism, it can never be performed again because the
> bot will recognize it.
on 12/31/08 1:15 PM, Ian Woollard at ian.woollard(a)gmail.com wrote:
There's an infinite number of ways to vandalise the wikipedia, and, in
practice, not all forms of vandalism can be detected by any known
design of bot, or humans with complete reliability for that matter.
I know something of machine learning myself, although I am not an
expert. In principle it can learn anything, in practice, there are
many problems and if you have *any* other way to do something, you're
normally better off.
Vandalism/spam is a difficult enough problem that *any* method should
be investigated and if it is found to be effective, applied, not
simply technological ones. But we need to stick to proportionality- we
should never use a sledgehammer to crack a nut.
Jarlaxle is only 19; as I understand it the human brain does not fully
mature until maybe 25. Unless he's actually mentally ill (which is by
no means inconceivable) he is likely to stop of his own accord at some
point.
You are treading on dangerous and uncertain ground here, Ian: the difference
between mental health and emotional health. I am not suggesting that the guy
be dragged off in chains to somewhere. But a strong, in-person message -
both to him AND his parents - from an authority spelling out the
consequences if he does not stop could go a long way towards resolving this.
Marc Riddell