Hallo Martin, first what you should do is to subscribe to the DSPAM mailing list. If you don't subscribe then the mailing list software will tag your message as spam and not deliver it to the other list members. Only list admins will get a copy of your post and are able to respond (that's what I am doing now).
> Today I received a spam email, which dspam thought was innocent, > so I told dspam to retrain it like this: > > dspam --user martin --class=spam --source=inoculation < msg > > No matter how many times I run this, including with --source=error: > > dspam --user martin --class=spam --source=error < msg > > it still insists that the message is innocent: > > dspam --user martin --stdout --classify < msg > > X-DSPAM-Result: martin; result="Innocent"; class="Innocent"; > probability=0.0000; confidence=0.75; signature=N/A > > To add insult to injury, the "confidence" value keeps going up! > > dspam_stats -H gives this result: > > martin: > TP True Positives: 2740 > TN True Negatives: 1272 > FP False Positives: 0 > FN False Negatives: 7 > SC Spam Corpusfed: 1205 > NC Nonspam Corpusfed: 0 > TL Training Left: 1228 > SHR Spam Hit Rate 99.75% > HSR Ham Strike Rate: 0.00% > PPV Positive predictive value: 100.00% > OCA Overall Accuracy: 99.83% > > Am I missing something? > Yes! You need to send us more information: - DSPAM version used - content of dspam.conf - output you get when executing "dspam --version" - what OS/Distro? > -- > Martin -- Kind Regards from Switzerland, Stevan Bajić ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Dspam-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspam-user
