Quick question about spamoracle: Is it important to try to keep the
number of sample spams and sample non-spams roughly the same when
constructing the database?  (E.g., if I present it with 100 times as many
non-spam examples as spam examples, will that tend to make it very
conservative, or even too conservative?)

And another (or maybe it's the same one): When it correctly judges a
message to be spam, is it a good idea to add this message to the database
as a new spam example, to reinforce what happened?  

Thanks for a great tool, Xavier!

        Benjamin


-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners