SPAM Update

November 29th, 2006

HeadacheAs if you are interested in knowing more about spam…

But I am since it is turning into a nightmare for me. Here are my SPAM stats:

Filtered Mail
149 Good Messages
2.707 Spam Messages (95%)
1.346 Spam Messages Per Day

SpamSieve Accuracy
10 False Positives
5 False Negatives (33%)
99.5% Correct

Showing Statistics Since
27-11-06 15:37

Yes, this is just for 2 days! 1346 spam messages and only 149 real ones. How about that?

The main reason why I get all this SPAM is because I also check the postmaster- and info@fleck.com accounts. This domain has been collecting SPAM for years so when we bought that domain SPAM just started pouring in from day one.

Two days ago I bought a serious spam filter for Mac OS X Mail which I hope will save me. It does seem to work well but I’m still training it and 5 false negatives is way to much to be practical.

I hope it will do better in a few days…


6 Responses to “SPAM Update”

  1. jcvangent on November 29, 2006 5:48 pm (17:48)

    Maybe it is time to start using MailScanner with spamassassin, works like a charm here, especially with the FuzzOCR plugin, which scans the content of images for spam (handy with all the incoming spammails with gif images attached to it)…turned my clean my from 20% down till 10%, and the only false positves are maybe 5 spammails a day nowadays (with around 600 mails being processed every day)…could have saved you the money for the Mac OS X spam filter ;)

  2. Boris Veldhuijzen van Zanten on November 29, 2006 5:58 pm (17:58)

    I’m using SpamSieve which implements a Bayesian algorithm to “learn” what spam looks like. It is supposed to be even better than MailScanner with spamassassin and free for the first 30 days. I’ll decide in 28 if I want to go on with it…

  3. jcvangent on November 30, 2006 12:08 am (0:08)

    the only problem with bayesian is that those spam with gif images almost everytime gets through, because they don’t use random words in their mails but random sentences so bayesian is very hard to train with these mails…but if you configure it right MailScanner + spamassesin + FuzzyOCR is the way if you ask me, talk to you in 28 days ;)

  4. Johannes de Jong on November 30, 2006 7:49 am (7:49)

    Boris,

    With SPAM figures like that I’m amazed you replied so fast to my emails.

    Thanks

    Johannes

  5. Uri L. on December 4, 2006 1:23 am (1:23)

    today I deleted 4 email accounts I’ve been using, cause I couldn’t handle the spam anymore. It wasn’t at the level of 1000’s per day, but I missed too many important messages over time because of this, not to mention how outlook hangs the computer .

    Btw, why not installing server-side spam blocker?

  6. The Internet Entrepreneur » My Ongoing Fight Against Spam on January 5, 2007 12:07 am (0:07)

    [...] It is time for another SPAM update! I posted the first one on November 29, 2006. At that moment I received an average of 1.346 spam messages per day and I just started using SpamSieve, for two days, to filter out all of it. [...]

Trackback URI | Comments RSS

Leave a Reply

Name

Email

Website

Speak your mind

retaggr