[Spambayes-checkins] website background.ht,1.19,1.20
Tony Meyer
anadelonbrin at users.sourceforge.net
Mon Aug 9 08:16:35 CEST 2004
Update of /cvsroot/spambayes/website
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv30459
Modified Files:
background.ht
Log Message:
Add a link to Gary's Linux Journal paper.
Add a link to my (and Brendon's) CEAS'04 paper.
Index: background.ht
===================================================================
RCS file: /cvsroot/spambayes/website/background.ht,v
retrieving revision 1.19
retrieving revision 1.20
diff -C2 -d -r1.19 -r1.20
*** background.ht 9 Jul 2004 00:37:25 -0000 1.19
--- background.ht 9 Aug 2004 06:16:32 -0000 1.20
***************
*** 13,19 ****
<a href="http://radio.weblogs.com/0101454/stories/2002/09/16/spamDetection.html">interesting essay</a>
suggesting some improvements to Graham's original approach.</li>
! <li>Gary also wrote a <a href="http://www.linuxjournal.com">Linux Journal</a>
! <a href="http://www.linuxjournal.com/article.php?sid=6467">article</a>
! about this.</li>
</ul>
<p><i>more links? mail
--- 13,18 ----
<a href="http://radio.weblogs.com/0101454/stories/2002/09/16/spamDetection.html">interesting essay</a>
suggesting some improvements to Graham's original approach.</li>
! <li><a href="http://www.linuxjournal.com/article.php?sid=6467">Gary Robinson's
! Linux Journal article</a> discussed using the chi squared distribution.</li>
</ul>
<p><i>more links? mail
***************
*** 265,266 ****
--- 264,278 ----
</ul>
+ <h2>Papers about SpamBayes</h2>
+ <p>Tony Meyer and Brendon Whateley wrote a
+ <http://www.ceas.cc/papers-2004/136.pdf>paper introducing Spambayes</a> for
+ the 2004 <a href="http://ceas.cc">Conference on Email and Spam</a> (CEAS 04).
+ The aim of the paper was to introduce SpamBayes (and so provide a paper to
+ reference for future work, allowing other authors to skip past the basic
+ stuff), and to introduce some of the main concepts that SpamBayes is based
+ on (like the importance of the 'unsure' range). Limited results from
+ testing the unigrams/bigrams tiling experiemental option, and from various
+ training regimes are also included. The intent was <strong>not</strong>
+ to demonstrate the superioriy of SpamBayes. Note that a great deal of the
+ background information about the history of SpamBayes is sourced from this
+ page.</p>
More information about the Spambayes-checkins
mailing list