Matus UHLAR - fantomas wrote:
>
> On 29.10.07 08:18, tad1214 wrote:
>> I have bayes set to auto learn on a catchall account, which I am now
>> thinking
>> is a bad Idea, and now everything that is sent is pretty much flagged as
>> 99-100% Chance spam by bayes. Usually that is it, but sometimes there are
>> little other things that are getting FP's. I am fairly new to Spam
>> Assassin
>> so please forgive me if I make mistakes.
>
>> So how do I fix this issue with nearly everything being flagged as spam.
>> Pretty much 100% of the mail being delivered to catchall is spam, but I
>> think that SA is starting to think even common words are bad.
>
> feed some hams into SA too. If there are just a few hams and many spams,
> every word can look spammy
> --
> Matus UHLAR - fantomas, [EMAIL PROTECTED] ; http://www.fantomas.sk/
> Warning: I wish NOT to receive e-mail advertising to this address.
> Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
> Windows found: (R)emove, (E)rase, (D)elete
>
>
Ok so I just threw a few hundred (thousand?) hams at it, we will see if that
helps, here is my dump magic
Pine# sa-learn --dump magic
0.000 0 3 0 non-token data: bayes db version
0.000 0 110832 0 non-token data: nspam
0.000 0 11160 0 non-token data: nham
0.000 0 546924 0 non-token data: ntokens
0.000 0 1187210461 0 non-token data: oldest atime
0.000 0 1193678380 0 non-token data: newest atime
0.000 0 1193678063 0 non-token data: last journal sync
atime
0.000 0 1193656666 0 non-token data: last expiry atime
0.000 0 124511 0 non-token data: last expire atime
delta
0.000 0 61843 0 non-token data: last expire
reduction count
--
View this message in context:
http://www.nabble.com/Everything-beying-flagged-positive-by-Bayes-tf4712332.html#a13472281
Sent from the SpamAssassin - Users mailing list archive at Nabble.com.