On Tue, 1 Mar 2011 23:42:22 -0800
Ron Garret <r...@flownet.com> wrote:

> A fourth question: I start with an empty database.  I take a single message 
> and feed it in with:
> 
> dspam --user foo --class=spam --source=corpus < testmsg
> 
> I look in the database and everything looks good: a bunch of spam tokens have 
> been created.  But then I ask it to classify the same message that I just 
> told it was spam:
> 
> dspam --user foo --classify < testmsg
> 
> and the result is:
> 
> X-DSPAM-Result: foo; result="Innocent"; class="Innocent"; probability=0.0023; 
> confidence=1.00; signature=N/A
> 
> Perhaps not coincidentally, this is the exact same result that I get when I 
> run --classify against a completely empty DB.
> 
> What am I doing wrong?
> 
Read http://www.mail-archive.com/dspam-user@lists.sourceforge.net/msg01436.html 
(especially the stuff regarding TL).


> 
> On Mar 1, 2011, at 1:34 PM, Ron Garret wrote:
> 
> > I'm trying to get a deep understanding of how DSPAM works so I'm running it 
> > manually from the command line.  I've gotten to the point where this works:
> > 
> > [ron@ip-10-112-7-132:~]$ sudo dspam --user foo --classify < testmsg
> > X-DSPAM-Result: foo; result="Spam"; class="Spam"; probability=1.0000; 
> > confidence=0.00; signature=N/A
> > 
> > Question 1: Is this the right way to classify a message?  dspam 
> > --deliver==summary seems to do the same thing as dspam --classify except 
> > that --deliver-summary also produces a signature.  Is that right?  Are 
> > these commands more or less equivalent?
> > 
> > Question 2: What is the difference between RESULT and CLASS?
> > 
> > Question 3: What is the difference between PROBABILITY and CONFIDENCE?
> > 
> > Thanks,
> > rg
> > 
> 
> 
> ------------------------------------------------------------------------------
> Free Software Download: Index, Search & Analyze Logs and other IT data in 
> Real-Time with Splunk. Collect, index and harness all the fast moving IT data 
> generated by your applications, servers and devices whether physical, virtual
> or in the cloud. Deliver compliance at lower cost and gain new business 
> insights. http://p.sf.net/sfu/splunk-dev2dev 
> _______________________________________________
> Dspam-user mailing list
> Dspam-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/dspam-user
> 

------------------------------------------------------------------------------
Free Software Download: Index, Search & Analyze Logs and other IT data in 
Real-Time with Splunk. Collect, index and harness all the fast moving IT data 
generated by your applications, servers and devices whether physical, virtual
or in the cloud. Deliver compliance at lower cost and gain new business 
insights. http://p.sf.net/sfu/splunk-dev2dev 
_______________________________________________
Dspam-user mailing list
Dspam-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/dspam-user

Reply via email to