Foxmail anti spam rules

[PerthMETRO.net]

Does anyone know any more on what foxmail has based it's anti spam
rules and its different levels? TIA Peter

 

[rm]

Le Thu, 29 Apr 2004 21:59:00 +0800, PerthMETRO.net a écrit :

Does anyone know any more on what foxmail has based it's anti spam
rules and its different levels? TIA Peter

Foxmail use white-list to by-pass others spam-filters and black-list (not
very usefull for spam-fighting) and internals rules (fixed and not
documented) and a very good bayesian filter (inevitably good... because it
needs 1000 spams AND 1000 no-spam messages to be learned before being
activated!)
rules and bayesian filters can be ajusted on three level each.
can't tell you more... it's vey efficent but need to download all message,
so not very good for bandwidth ;-)
another "solution" is to kill spam message with native remote POP3 server
pre-viewing feature (dl only mail headers.)

@+

 

[PerthMETRO.net]

Le Thu, 29 Apr 2004 21:59:00 +0800, PerthMETRO.net a écrit :


Foxmail use white-list to by-pass others spam-filters and black-list
(not very usefull for spam-fighting) and internals rules (fixed and
not documented) and a very good bayesian filter (inevitably good...
because it needs 1000 spams AND 1000 no-spam messages to be learned
before being activated!)
rules and bayesian filters can be ajusted on three level each.
can't tell you more... it's vey efficent but need to download all
message, so not very good for bandwidth ;-)
another "solution" is to kill spam message with native remote POP3
server pre-viewing feature (dl only mail headers.)

@+

Hi rm, what I really wanted to know was how foxmail made up its RULES
per say.... not its bayesian/blacklist/whitelist etc..... Im assuming
its some kind of RegEx filter. I just would like to know what
determines the low medium hign settings?

 

[rm]

Le Sat, 1 May 2004 09:02:30 +0800, PerthMETRO.net a écrit :

Hi rm, what I really wanted to know was how foxmail made up its RULES
per say.... not its bayesian/blacklist/whitelist etc..... Im assuming
its some kind of RegEx filter. I just would like to know what
determines the low medium hign settings?

i just know that internals rules are based on words's list and perhaps some
typical headers and significant mail structure...
the file Antispam.dll is used for this filter (bad words are "visible"
inside it)

@+