hi all
i am a newbie in the dspam area
i was able to install dspam and get it working by following instructions from a forum. But i still am confused in many areas
my system is fedora 4, qmailtoaster - qmail, vpopmail, chkuser, simscan, clam and dspam
this is how it is currently working
data stored in
/etc/dspam/data/testdomain.com/user1
/etc/dspam/data/testdomain1.com/user1
i have configured it for a single domain : testdomain.com
# cd /home/vpopmail/domains/testdomain.com
# edit .qmail-default file
.qmail-default
# edit last line
| /etc/dspam/bin/dspam --deliver=innocent --user $EXT@$USER --debug --mode=teft --feature=chained,noise,whitelist --stdout -p -m | /home/vpopmail/bin/vdelivermail '' bounce-no-mailbox
# save and exit
For training dspam i created two email ids spam@testdomain.com and notspam@testdomain.com and placed two files in the testdomain.com directory as below
file .qmail-spam with statement
|/etc/dspam/bin/dspam --user $NEWSENDER --debug --mode=teft --class=spam --source=error
file .qmail-notspam with statement
|/etc/dspam/bin/dspam --user $NEWSENDER --debug --mode=teft --class=innocent --source=error --stdout
the above setup is working dspam is training itself but it is slow to learn.
now I wish to proceed ahead and provide out-of-box spam filtering for all users by training dspam in advance.
I am confused by the documentation and the instructions provided on forums and readme file
I need is specific instructions assuming that i am a total newbie.
i am using qmailtoaster -- qmail + vpopmail
what i need
1) i wish to configure dspam with groupname:classification:*globaluser to offer immediate spam filtering for all users
########
groupname:classification:*globaluser
This will automatically add globaluser as a classification peer to all users. Any user who has less than 1000 innocent messages or 250 spam messages in their corpus, or whose filter is uncertain about a particular message will consult the global dictionary for an answer.
Global groups will need to be trained using corpus or other means, or by using the dspam_merge tool. the global user (in this case 'globaluser') is treated just as any other user on the system.
########
I can collect 1000s of spam mails and ham mails into two seperate email ids spamcorpus@testdomain.com and hamcorpus@testdomain.com for training but how do we train dspam
ie how do we inform dspam the location of this corpus. which command do we provide the path to dspam
is there any training file for this ?
2) similarly i wish to setup up innoculation box also -- ie circulate a honeypot email user innoculate@testdomain.com and spread it around so that it is picked by spammers.
i need the method by dspam can be trained on such spam mails also
pl provide the exact step by step instructions for both of the above
thanks in advance for all help
Regds
Rajesh Mahadevan
##############
Open source documentation is as important as development. For every hour spent on development a thousand is lost in figuring out how it works -- unless well documented.
Received on Mon Mar 12 21:14:58 2007
This archive was generated by hypermail 2.1.8 : Tue Mar 13 2007 - 00:00:06 CET