Update of /cvsroot/mahogany/M/doc
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv24223/doc
Modified Files:
Manual.htex
Log Message:
added spam filters section, briefly described dspam
Index: Manual.htex
===================================================================
RCS file: /cvsroot/mahogany/M/doc/Manual.htex,v
retrieving revision 1.105
retrieving revision 1.106
diff -b -u -2 -r1.105 -r1.106
--- Manual.htex 11 Jul 2004 01:00:05 -0000 1.105
+++ Manual.htex 13 Jul 2004 14:51:15 -0000 1.106
@@ -32,5 +32,5 @@
\vfill{}
\title{{\LARGE \textsl{Mahogany User Manual}}\\
-{\Large Version 0.67 ``???''}}
+{\Large Version 0.67 ``Constance''}}
\vfill{}
@@ -81,4 +81,5 @@
\begin{itemize}
+ \item New integrated statistical spam filter (DSPAM)
\item Folders can now be dragged in the folder tree to reorder them
\item New \texttt{--import} command line option to simplify sharing
@@ -88,6 +89,22 @@
\item New \texttt{nop()} (stop filter processing) filter action added
\item The "Set From from To" option (\ref{IdentityPage}) is more useful
+ \item When replying to PGP-encrypted messages, clear text is used for
+ replies and not the original message.
\end{itemize}
+By far the main new feature in this version is the integration of DSPAM
+According to http://www.nuclearelephant.com/projects/dspam/:
+
+\begin{quote}
+DSPAM (as in De-Spam) is an extremely scalable, open-source statistical
+anti-spam filter. While most commercial solutions only provide a mere 95\%
+accuracy (1 error in 20), a majority of DSPAM users frequently see between
+99.95\% (1 error in 2000) all the way up to 99.991\% (2 errors in 22,786).
+\end{quote}
+
+Please see \ref{DSPAM} section for the description of how to configure and use
+DSPAM.
+
+
\subsubsection{0.66 against 0.65}
@@ -1535,4 +1552,61 @@
+\subsection{Spam filtering}\label{spam}
+
+Mahogany supports pluggable spam filter modules. In theory it means that
+anybody can write such modules but in practice only two of them currently exist
+and are described in details below.
+
+However independently of the spam filters used, the interface to them remains
+always the same:
+\begin{itemize}
+ \item The spam filter options can be edited in
+ \MenuCmd{Edit|Spam filters...} dialog. Mostly you can rely on the
+ default values (but see \ref{DSPAM}).
+ \item You can check whether the currently selected messages are deemed to
+ be spam by using \MenuCmd{Message|Spam|Check message...} command.
+ This is mainly useful for testing, see below for how to check for spam
+ automatically.
+ \item If a statistical filter mistakenly recognizes a message as spam (so
+ called false positive, this should hopefully be very rare) or misses a
+ spam message, you can correct its mistake by using the two other
+ commands of \MenuCmd{Message|Spam} submenu. Please note that you
+ should \emph{not} use these commands for the messages not seen by the
+ spam filter at all, i.e. they don't work for training it but only for
+ correcting its errors.
+\end{itemize}
+
+To actually use the spam filters you need to configure a filter rule (see
+\ref{Filters}) using them. Usually you just create a rule named "Spam" using
+the "Seems to be spam" test and "Move to some Junk folder" (you will need to
+create this junk folder too) as its action. If you feel very confident in the
+spam filters (or just have backup of your mail somewhere) you can also
+configure the rule to delete the messages recognized as spam immediately but
+this is not recommended.
+
+
+\subsubsection{Simple header analyzer}
+
+This is a very simple filter which looks for some common characteristics of
+spam mail in the message headers. Its accuracy is much worse than that of DSPAM
+but it is also much faster to execute and doesn't need any training. You may
+configure its options in the \MenuCmd{Edit|Spam filters...} dialog.
+
+\subsubsection{DSPAM}\label{DSPAM}
+
+DSPAM is a statistical spam filter which means that to use it in the optimal
+way you need to \emph{train} it first. Just open the \MenuCmd{Edit|Spam
+filters...} dialog and use the \texttt{Train} button on the DSPAM page. For the
+best results you should train the filter with 2500 spam and 2500 non-spam (ham)
+messages. Be warned that this can take a rather long time and also consumes
+significant amount of disk space (of order of 50Mb) -- however the results are
+well worth it.
+
+If you don't have enough messages to train it with, you can start using it
+immediately but in this case you should be ready to manually mark messages as
+spams as explained in \ref{spam}. Please also remember to actually create a
+spam filter rule to really use DSPAM at all.
+
+
\section{Sending Mail}
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Mahogany-cvsupdates mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mahogany-cvsupdates
