Is there a way to learn what bayes learned so far?
> On Oct 27, 2015, at 4:35 PM, John Hardin <jhar...@impsec.org> wrote: > >> On Tue, 27 Oct 2015, j...@lexoncom.com wrote: >> >> example mail sa headers: > > Is this from a spam? > >> X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on >> ip-10-254-37-89.us-west-2.compute.internal >> X-Spam-Level: *** >> X-Spam-Status: No, score=3.1 required=5.0 tests=BAYES_00,HTML_MESSAGE, > > BAYES_00. You *do* have ham and spam trained, and bayes *is* in use. > > If this is a spam, your Bayes appears to be mistrained. That might explain > why so many spams are getting through. > > If you have autolearn turned on, turn it off. > > Collect hand-classified corpora of several hundred hams and several hundred > spams, then wipe and retrain your Bayes. > > If your userbase is small enough to collect and train on just misclassified > messages, then leave autolearn turned off and just train misclassifications > and messages that don't hit either BAYES_00 or BAYES_99. > > -- > John Hardin KA7OHZ http://www.impsec.org/~jhardin/ > jhar...@impsec.org FALaholic #11174 pgpk -a jhar...@impsec.org > key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79 > ----------------------------------------------------------------------- > ...the Fates notice those who buy chainsaws... > -- www.darwinawards.com > ----------------------------------------------------------------------- > 4 days until Halloween
