[Spambayes] Korean Spam
tameyer at ihug.co.nz
Tue Dec 13 11:49:28 CET 2005
> I get many Korean emails. Will your product reject Korean characters?
Do you get many Korean spam, that you want deleted, or many Korean
ham, that you don't want deleted, or both Korean ham and spam?
SpamBayes doesn't know any languages, so in theory should work no
differently: *except* it does assume that words are separated by
whitespace. I don't know if this is the case with Korean or not; I
suspect it may not be.
(If the problem is getting rid of Korean spam, then that's easy, and
there's an additional option you can turn on to make it quicker. If
the issue is Korean ham, but there's no Korean spam, then that's easy
too, because the ham won't look anything like any of the spam. I'll
assume that you get a mix).
I've heard that SpamBayes works fine (with minor, mostly interface,
changes) for Japanese. My very limited knowledge of Asian languages
leads me to think that this will be similar with Korean. However,
there really hasn't been anything in the way of testing, so it's not
There is a large advantage to trying SpamBayes: it's free. Download
1.0.4 and install it, and see how it goes. You set up your own
rules, and still get all your mail, so it's basically risk free. If
it works, then all is good (it'd be great to get feedback), and if it
doesn't, you can either ask for suggestions (the "clues" for a sample
misclassified message would be great) or switch to something else
(with a full refund <wink>).
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.
More information about the SpamBayes