Home | Contact Us | FAQ | Search & Site Map | Link to Us
Sign In | Join | Other 45 Sites in Network
Home
Discussion Groups
General
GeneralPortable MacsHardwareNetworking
Applications
Mac ApplicationsEudoraFirefox / MozillaInternet ExplorerOutlook ExpressMS OfficeEntourageExcelPowerPointWordVirtual PCMedia PlayerOther MS Products
Programming
Mac ProgrammingCodeWarriorPerl
Country Specific
Australian Mac GroupUK Mac Group

Mac Forum / General / Networking / July 2005



Tip: Looking for answers? Try searching our database.

Mac Mail junk filter not working

Thread view: 
Enable EMail Alerts  Start New Thread
Thread rating: 
Amanda Robin - 12 Jul 2005 22:20 GMT
I'm running 10.4.1 on an iMac G5. I migrated from an older version of
Eudora to Mac Mail, using the Eudora Mailbox Cleaner application. I used
only 1 filter in Eudora, and I have since deleted it.

My Junk mail filter in Mail is not marking any junk, of which I get
plenty.

I have it set in Training mode.

My settings are:
Enable junk mail filtering, checked
Leave it in my Inbox, checked
Exempt if sender of message is in my address book

I have tried "Trust junk mail headers set by my ISP" both enabled and
disabled.

When I click on Advanced, the filter description looks right (If mail is
Junk AND if it is not from someone in my address book, color it brown).

Just for fun, after this was obviously not working, I added the same
filter under my Rules setting and made it first in line. No help.

I get 50+ pieces of spam a day.

So for the last week I have been dutifully marking them as junk, but I
have yet to have the system identify an incoming one as such.

Is this a Tiger problem? I know that when I first tried the Mac Mail
program, junk filtering was working. But I only started seriously using
it after I installed Tiger, and now it isn't.

Thanks for any help,

Amanda
Marc Heusser - 13 Jul 2005 06:24 GMT
In article
<amanderr-C4C9EF.16204812072005@newssvr11-ext.news.prodigy.com>,

> I'm running 10.4.1 on an iMac G5. I migrated from an older version of
> Eudora to Mac Mail, using the Eudora Mailbox Cleaner application. I used
> only 1 filter in Eudora, and I have since deleted it.
>
> My Junk mail filter in Mail is not marking any junk, of which I get
> plenty.
...
> Is this a Tiger problem? I know that when I first tried the Mac Mail
> program, junk filtering was working. But I only started seriously using
> it after I installed Tiger, and now it isn't.

Can't help you specifically with Mail, but in just in case you do not
get it filtering, try SpamSieve in conjunction with Mail.
Works well for me (using Eudora), and I do get more junk than you
unfortunately.

HTH

Marc

Signature

Switzerland/Europe
<http://www.heusser.com>
remove CHEERS and from MERCIAL to get valid e-mail

Amanda Robin - 13 Jul 2005 22:20 GMT
In article <marc.heusser-3FB8E4.07242213072005@news.cybercity.ch>,
Marc Heusser <marc.heusser@CHEERSheusser.comMERCIALSPAMMERS.invalid>
wrote:

> In article
> <amanderr-C4C9EF.16204812072005@newssvr11-ext.news.prodigy.com>,
[quoted text clipped - 18 lines]
>
> Marc

Thanks to everyone who posted. Today, for the first time, it marked TWO
pieces as spam. Gosh, are spammers getting that smart, or is the
filtering that poor?

I use SpamCop for my other accounts, which does well, but the one in
question (a university account) I haven't figured out how to forward to
them.

I'll look into Spam Sieve, and also k e e p  o n  t r a i n i n g the
Mail program.

Thanks,
Amanda
Kyle Jones - 14 Jul 2005 06:14 GMT
>  Marc Heusser <marc.heusser@CHEERSheusser.comMERCIALSPAMMERS.invalid>
>  wrote:
[quoted text clipped - 25 lines]
> pieces as spam. Gosh, are spammers getting that smart, or is the
> filtering that poor?

The spammers are definitely getting better at evading Bayesian type
filters, a type which I think Apple Mail uses.  Also it takes a
while to train such filters.  I used a corpus of 13,000+ messages to
train my filter originally and have probably fed 150,000+ messages
through it since then.  And still the bletcherous "Confidence is
back" Viagra spam gets through.  Daily.  You won't get perfection,
but as you train the filter you should see rejection rates at or
above 90%.
Stefan Schwetschke - 15 Jul 2005 15:12 GMT
Amanda Robin schrieb:
[...]
> Thanks to everyone who posted. Today, for the first time, it marked TWO
> pieces as spam. Gosh, are spammers getting that smart, or is the
> filtering that poor?
[...]

Some of the spammers are astounishing smart. They have access to the
commonly used spam detectors, flood them with spam and look what remains
undetected. Then they optimise their spam to fool the filters. Only spam
that cannot be detected by the state-of-the art filters is spread to the
poor victims. I already saw a job advertisment for a position as spam
optimiser. Fortunately most of the spammers don't invest so much efford
and send out unoptimised spam. How much your spam filter can protect
you, depends on the smartness of the spamers that are after you.

In the good old days, people who were smart enough to have a spam filter
were also smart enough to ignore spam. Today, spam filters are too
convinient. There are people behind spam filters who would actually
respond to spam or even buy products advertised by spam. So it became
profitable too trick spam filters.

In other words: the more easy a spam filter is too handle, the less it
will protect you in the long run.

Geggo
Marc Heusser - 16 Jul 2005 12:33 GMT
> Some of the spammers are astounishing smart. They have access to the
> commonly used spam detectors, flood them with spam and look what remains
> undetected. Then they optimise their spam to fool the filters. Only spam
> that cannot be detected by the state-of-the art filters is spread to the
> poor victims.

That is why server side spam filters only filter part of the spam.
A true Bayesian filter on the client side (like SpamSieve) will take
care of the rest. And it cannot be circumvented as easily since it works
by true probabilities of your (individual) good mail and spam mail.
Therefore the spam filters will work on different parts according to
each individual's mails.
Get a good client spam filter, and train it. I get some 100-200 spam
mails a day, and SpamSieve takes good care of them :-)

HTH

Marc

Signature

Switzerland/Europe
<http://www.heusser.com>
remove CHEERS and from MERCIAL to get valid e-mail

Stefan Schwetschke - 18 Jul 2005 11:33 GMT
Marc Heusser schrieb:
[...]
> A true Bayesian filter on the client side (like SpamSieve) will take
> care of the rest. And it cannot be circumvented as easily since it works
> by true probabilities of your (individual) good mail and spam mail.
[...]

In theory your right. In practice, we all get the same spam. Of course,
the ham is different. But most of us would classify a Wikipedia article
as ham. If I was a spammer, I would train a Bayesian filter with real
spam as spam and Wikipedia articles as ham. Then I would optimise my
spam to pass that filter. I'm sure such a spam mail would fool most well
trained Bayesian filters. The trick is that most filters are in fact
very similar because they were trained with similar content for a
similar purpose. Of course, I'm not apamer. Even if I was a spamer, it
would not make sense to fool client side filtering. People smart enough
to use such a filter won't read spam mails.

Geggo
matchbox - 19 Jul 2005 18:38 GMT
I am not convinced by the argument that Spammers are getting better a
the explanation for this problem.   I was running Panther on a G5 an
my Mail
filtered out >99% of the junk.  I upgraded to Tiger and my Mail filter
out
nothing.  NOTHING.   Even after a month of training.   I don't see ho
the spammers could improve so much in that one day it took me t
convert
to Tiger, or even during that one month of training after I upgraded

--
matchbo
macosx.com - The Answer to Mac Support - http://www.macosx.co
Elliott Roper - 19 Jul 2005 19:56 GMT
> I am not convinced by the argument that Spammers are getting better
> as the explanation for this problem.   I was running Panther on a G5
[quoted text clipped - 3 lines]
> that one day it took me to convert to Tiger, or even during that one
> month of training after I upgraded.

Me too. I was playing silly buggers with the .plist to get PGP working
so I kept mousy quiet about the junk filter.

SpamSieve is working quite well. Next time I reboot I gotta pay. It
will be worth the money.

Mail is more than a bit flaky in 10.4.2 and 10.4.1. Don't do anything
too quickly while navigating your mailboxes with the keyboard.

Too many of the Mac's new wonder programs are playing fast and loose
with the HIG. Often in different ways. Somebody should bang Mail and
Safari and Finder programmers' heads together. Firmly!

Signature

To de-mung my e-mail address:- fsnospam$elliott$$
PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248

Elliott Roper - 13 Jul 2005 09:25 GMT
In article
<amanderr-C4C9EF.16204812072005@newssvr11-ext.news.prodigy.com>, Amanda
Robin <amanderr@yahoo.com> wrote:

> I'm running 10.4.1 on an iMac G5. I migrated from an older version of
> Eudora to Mac Mail, using the Eudora Mailbox Cleaner application. I used
[quoted text clipped - 29 lines]
>
> Thanks for any help,

I see this too. In my case, I suspect it was caused by me fooling with
the bundles settings in mail's plist in order to get an early version
of PGP-9 encryption working the way I wanted it to. Now I can't get
mail to resume filtering junk, even when I undo the changes I thought I
made.

I'm test-driving spam-sieve as a work around. It seems pretty good so
far.

Signature

To de-mung my e-mail address:- fsnospam$elliott$$
PGP Fingerprint: 1A96 3CF7 637F 896B C810  E199 7E5C A9E4 8E59 E248

DJ Craig - 13 Jul 2005 15:09 GMT
It seems like some people have great results with the Apple junk mail
filter, and for others it doesn't help at all.  It works great for me.
I use Mail to check 6 email addresses that are all posted online in
many places and get tons of spam, and the Mail junk filter gets rid of
all of it.  Maybe you should just keep on trying longer in training
mode.  I don't really know how long it took for mine to get trained,
since when I started using it I used to get very little spam, perhaps
one a day compared to around 100 a day now.
 
Sign In
Join
My Latest Posts
My Monitored Threads
My Blog
My Photo Gallery
My Profile
My Homepage

Start New Thread
Enable EMail Alerts
Rate this Thread



©2008 Advenet LLC   Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.