Re: A few questions about ifile features compared to other classifiers

Xavier DUTOIT <[email protected]> wrote on 22/09/2003 (18:11) :
> 2) Mail parsing . <br>
> Features like recognition and decoding of MIME attachments in
> quoted-printable and base64 encoding, Ignores HTML tags in emails,
> handling things like V'I'A'G'R'A (random choosed example ;), Scores
> only the Received, Subject, To, From, and Cc headers...<br>
> <br>

Well Ifile filtered your mail to my Ifile folder evenif you posted with
HTML!!!!

Please stop doing that to mailinglists. Before ifile I had a filter
which sent any mail in HTML to Spam.

> 3) A last thing about sort accuracy.<br>
> I read in one page that some of you reached 96% accurate
> classification. How have you calculated that ? Do you all have such
> high percentage ?<br>

Well. I have no numbers, but ifile filters about 10 mailinglists for me.
I get about 50-60 spams a day and most find their way to the Spam
folder. Perhaps 2-3 slip through to one of the mailinglists.

The only thing I have in my procmail is:

:0 fBw : $MAILDIR/.proc_lock
| $HOME/bin/ifile-tag-mail.sh $MAILDIR

:0:
* ^X-Ifile-Hint: \/.+
$MATCH

where ifile-tag-mail.sh lets ifile decide which mail folder the mail is
belonging to.

I run a script regularily which simply check each mail with ifile to see
which folder ifile think it should be in. If the folder is different
from the one ifile thinks it should be in I tell ifile to relearn the
mail to the new folder.

Just try it out.

PS: Ifile need a few days of training to filter the mailinglists
properly. But if you use mutt it is easy.

Preben