Hello neurowerx, Tuesday, July 15, 2003, 2:32:42 PM, you wrote:
nwd> 15-Jul-2003 21:16, [EMAIL PROTECTED] wrote: >> I don't know about other Bayesian filters but for POPFile, for example >> IP addresses are just "words" that it can use to classify messages. And >> it sure does learn and use IP lists. nwd> Good point. However, I believe that DNS blacklists are updated faster than nwd> popfile will "learn" IP addresses when you teach them manually. The problem with 'black lists' is that inevitably there are a lot of people on them that should not be. I maintain a list server and I can tell you that several of us (I talk with other list managers) get black listed every other month or so for various reasons, but mostly because the list maintainers do not do "due diligence" when someone says they got spamed from such and so address. So I will never use a black list to determine spam. So far I am still trying to get something bayesian that will work for me. I have hopes for the new bayesian plug-in for TB. -- Best regards, MikeD mailto:[EMAIL PROTECTED] Using The Bat! v1.63 Beta/7 on Windows ME 4.90 Build 3000 ________________________________________________ Current version is 1.62r | "Using TBUDL" information: http://www.silverstones.com/thebat/TBUDLInfo.html