Discussion: design of spam filter in frost

Hello,

a user on a frost board raised the idea to implement a Baysian learning
spam filter into frost. I think this idea is great.

The user design of such a filter could be similar to that what Mozilla
Thunderbird implemented: a global JUNK folder where all messages are
stored that don't pass the filter. The original target board is shown in
the junk folder, and the user can mark messages as NOT JUNK (message is
moved to original folder as new message), and he can mark messages in
normal folders as SPAM (message is moved to JUNK folder).

Once I found a nice open source Baysian filter library (does anyone know
a good one?) I could implement this into frost.

What are your opinions about this design? Any other ideas? Does someone
want to help me with this?

I just found this library: http://bayesjunktool.mozdev.org/index.html
I now will check if this is really a library that can be used in frost,
if this component is still supported and developed ... and maybe there
are better libraries. Please help me to find a good and supported library :)

rgds, bback.


-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf