Sorry if this has been discussed in the past...

I'd like to bring up the subject of the scoring of different rules which 
have, IMHO, a very high likelyhood of being spam but are scored rather 
low.  And I noticed a somewhat related fact, that there is no one rule 
that in itself decides something IS spam, no matter how incredibly likely 
it is. In other words, there exists no rule with a score over 5 points.

Some of those examples I included below. Some checks get scored 
unreasonably low in listing 1, and some that -in my opinion- deserve to 
get scored higher than 5 points, in listing 2.  
And this is only a very small subset.

Of course this is open to debate, but then again that's all I want; 
possibly a debate about how accurate the scoring is right now...

List 1:
score ALL_CAP_PORN 0.650 0.669 0 0
score PENIS_ENLARGE2 0.500 0.590 0 0.501
score UPPERCASE_50_75 0.794 1.137 0 0
score V+AG+A_ONLINE 1.100 1.101 3.151 4.056

If it were up to me, I'd say that giving only half a point to a mail that 
scores PENIS_ENLARGE2 is...  well, ludicrous.  Let's not kid ourselves. 
IF there are people who participate on a genuine mailinglist that 
discusses penis enlargement, let the burden fall on them to put those 
adresses in their whitelist, not the reverse.

List 2:
score SUBJ_V+AG+A 2.535 2.816 4.095 4.100
score INVALUABLE_MARKETING 2.900 2.800 1.119 0.701

I have problems with these scores. They are high, sure, but why isn't 
SUBJ_V+AG+A in itself, all by itself enough to score, say, six points ?
(or twenty for my part). Is there really ANYONE who believes that a mail 
with such a subject could in fact be something other than spam ? 

Just being curious at how this system came about...

Now of course the problem is getting this mail to the list. Cause with all 
these words and phrases in it, it's sure to register as spam, right ? ;-|
( In fact I had to rewrite the V-word for it to get through sourceforge )

Regards,
Maarten

-- 
This email has been scanned for the presence of computer viruses.

Maarten J. H. van den Berg   ~~//~~   network administrator
VBVB  -  Amsterdam  -  The Netherlands  -  http://vbvb.nl  
T +31204233288   F +31204233286   G +31651994273




-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?   SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk

Reply via email to