[Spambayes] Spambayes and GMail

[Tony Meyer]

> Does anybody have any ideas on how with some work we
> can either get Google to implement SPAMBAYES as their filter,

This is extremely unlikely to happen.  Most people are pretty happy with
gmail's filter, so it's unlikely that they would be planning on changing it.
It's likely that they wrote their own filter, rather than using an existing
one, which means that they would be more likely to simply work on theirs.

(I met a researcher from Google, Mehran Sahami, last year, who wrote an
early paper on Bayesian email filtering - that's not what he works on now,
but with people like that there, I doubt they're using someone else's
system).

I'm sure that the people at Google that work on their filter keep up to date
with what others are doing, and try out ideas that work for others,
including spambayes.

> or get a SPAMBAYES on my machine to filter the GMail ?

Remotely filtering webmail isn't really practical, although it can be done
if the web interface is stable enough (it can simulate doing it manually).
This would be a lot of work, though, and given the level of satisfaction
with gmail's filter, I don't see it happening any time soon.

If you read your gmail mail via POP3 in another mail client, then you can
use the spambayes POP3 proxy to filter mail (although I don't believe you
can get gmail to send mail it thinks is spam out through POP3, which leaves
any gmail fp's where they were).

> One problem is that there is no way to turn OFF GMail's 
> SPAM filter as of now,

If you want to be able to turn it off, then suggest that to them.  I've
heard good things about their suggestion system (that suggestions do
actually get read), so there's a chance it would happen.

> and unfortunately GMails SPAM filter has an unacceptable
> false positive rate ( >> SPAMBAYES == 0.000001 %).

No-one here would claim a fp rate that low for spambayes.  (I'm sure that
with the right training and right mix of incoming mail you could achieve it,
but it wouldn't be an average figure).

> Does any SPAMBAYES developer work at Google? 

Not as far as I know, and I presume that I would.

> Anybody have ANY influence at Google?

Lots of people, I'm sure, but no-one that I know.

=Tony.Meyer

-- 
Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes.
http://www.massey.ac.nz/~tameyer/writing/reply_all.html explains this.