On Wed, Aug 26, 2009 at 2:37 PM, Karl Pielorz<[email protected]> wrote: > It'd be relatively trivial for us to take the mail sent by users - and run > it through training on dspam. > > On the surface it seems like a good idea - as mail going out would have > patterns in that you'd expect back at some point?
I've also thought about this as being a good idea. Returning mails often quote the original mail, and also contain sender/recipient, so a returning mail with these tokens should be a good indication of ham. But I have not tried it, yet... should be interesting to hear other comments on this. One issue could be skewing the ham/spam ratio, as no spam is going out. Also headers are different, which could give some misweight perhaps? BR Sven ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Dspam-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspam-user
