I have tried to use the sa-learn --spam/--ham feature with my mbox. I was told you need to do it when you have hundreds of examples of --ham and --spam. I have like 1000 spams in my assassinated folder and about 2000 good mails in my inbox. I used the sa-learn and i got back "learned from 1 file" ..
Does that mean that it only learned from 1 instance in that file or only from 1 files because of the way mbox is setup. Also i don't see any "sa-learn" instances in my spam tests. I have seen 1 in about 100 or so spam/hams. Maybe i'm just not being patient enough.
If you look in the "tools" subdirectory of the tarball there's a tool called check_bayes_db... if you run it, the top of the output will tell you how many spam/ham emails it's learned from.
a quick sample: 0.000 0 709 0 non-token data: nspam 0.000 0 57 0 non-token data: nham 0.000 0 34253 0 non-token data: ntokens 0.000 0 0 0 non-token data: oldest age 0.000 0 767 0 non-token data: current scan-count 0.000 0 0 0 non-token data: last expiry scan-count <snipped off a very long chunk of token data>
------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
