On 7 May 2007 at 12:27, David wrote: > The "reinventing the wheel" comment was just noting that Bayesian > filtering doesn't seem to be the sole focus of ASSP, but it is the sole > and primary focus of other projects. I can't find any information on how > ASSP handles it's bayesian filtering. I would really love to see some.
>From the website, DSPAM itself uses a variety of filtering methods - not just >Bayesian. (For ASSP's methods, check out the old documentation - links from the wiki IIIRC.) Among other issues would be the portability to Windows - as it's complied C and the Windows porting/packaging appears to have stopped. > I'm going to attempt to perform some testing between the two > engines.Using an identical spam/ham corpus, I hope to train both filters > to see which is more effective in catching spam. I promise no results, > as I'm hardly savvy at this sort of thing, but I'll do what I can. Let us know what you find. Although it's not easy to do direct comparisons between filtering packages, my experience running another Bayesian filter over 18 months is that ASSP is more accurate in both FPs and FNs. paul ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
