On Fri, 2003-12-26 at 02:00, schafer wrote: > [...] > Exhibit: > > ---- Start SpamAssassin results > 7.10 points, 5.5 required; > [...] > * 3.0 -- BODY: Bayesian classifier says spam probability is 99 to 100% > [score: 0.9988] > [...] > * -4.3 -- AWL: Auto-whitelist adjustment > ---- End of SpamAssassin results
Now I'm relatively new to spamassassin, but as I understand it, these results indicate that: 1. Whoever has configured Lenny's spamassassin (just curious: where is it running?) has trained the bayes filter to tag similar messages as spam (the bayesian classifier results). 2. A number of "good" messages have been received from the same list (the AWL adjustment), so most of the messages from Lenny's list have presumably been OK. So corrective action would be to un-train it on this particular message by feeding it to sa-learn with the --ham option. The spamassassin documentation has a lot of info on the importance of training properly to avoid problems. In any case, if somebody specifically trains a filter to flag your messages as spam, there's not much you can do about it, whether they're using spamassassin, or any off-the-shelf product. It looks like this has happened in this instance. I think the fix will really depend on WHO is running spamassassin in this case. - Bob ------------------------------------------------------- This SF.net email is sponsored by: IBM Linux Tutorials. Become an expert in LINUX or just sharpen your skills. Sign up for IBM's Free Linux Tutorials. Learn everything from the bash shell to sys admin. Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
