On 01/20/2016 10:28 PM, Quanah Gibson-Mount wrote:
--On Wednesday, January 20, 2016 4:26 PM -0500 Wrolf <wr...@wrolf.net>
wrote:
Is Marc's approach "novel" and "non-obvious"? (Patents must be novel,
non-obvious, and useful.)
I think plenty of people have supplied prior art, and that the concept
itself is obvious since other things implement similar ideas. I.e,. see
bogofilter, dspam, and bayes in general.
Agreed.
If anything SA could possibly make use of a plugin using some fuzzy
learning. Something in the line of Spamsum would be nice to have...
(https://www.samba.org/ftp/unpacked/junkcode/spamsum/README)
but it still isn't going to be a novel approach. Just one more....
