On Wed, 03 Dec 2014 10:39:41 +0000 Steve Freegard wrote: > > you would install that seriously on a sensible server? > > if a project is dead it should say that clear and close > > > > > > To add my voice to the others; I started using DSPAM fairly recently > and whilst I don't use the web interface (I cooked up my own methods > of training), I've found it to be completely stable and I've had > absolutely no issues with it at all and accuracy once trained has > been excellent. > > If you're willing to spend the time to learn it and get your hands > dirty a bit, then DSPAM is well worth the effort IMO. It's way > faster than SpamAssassin (due to the bazillion network tests that SA > does) and is a considerably more advanced implementation of Bayes > than SA.
I've scored Bogofilter (with multi-word tokenization ) and DSPAM (with various settings) into SpamAssassin for some years, and collectively it works well. However DSPAM does seem to be the weakest link. It's been consistently out-performed by Bogofilter in terms of both the TP and FP rate over the entire period. What I'm seeing is that DSPAM seems to have some blind-spots that let it down. There are some types of spam that it unaccountably struggles with that are caught easily by Bayes and Bogofilter. I'd be surprised if it isn't being let down by bugs. As always YMMV. ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ Dspam-user mailing list Dspam-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dspam-user
