[Dspam-user] makeCorpus=on

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[Dspam-user] makeCorpus=on

ML mail
Hi,

I just switched from TOE training mode to TUM in the hope of better results for new e-mail accounts. As recommended in the README I have turned on the makeCorpus preference.

Now regarding to that makeCorpus preference I was wondering what is the point exactly in this option? or how is it useful to DSPAM? The only thing I notice DSPAM doing is saving the mails in /var/spool/dspam/data/domain/user/corpus/spam and and nonspam directories. I run DSPAM with debugging on and so I see the "writing to corpus file ..." but if I understand it just writes the file and that's it. So again I don't get the point how these corpus files are useful to DSPAM? ;) Anyone can explain? I did not find any more infos in the README about that.

Thanks so much!
ML

------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Dspam-user] makeCorpus=on

Tom Hendrikx
On 12/13/2013 11:34 AM, ML mail wrote:

> Hi,
>
> I just switched from TOE training mode to TUM in the hope of better
> results for new e-mail accounts. As recommended in the README I have
> turned on the makeCorpus preference.
>
>
> Now regarding to that makeCorpus preference I was wondering what is
> the point exactly in this option? or how is it useful to DSPAM? The
> only thing I notice DSPAM doing is saving the mails in
> /var/spool/dspam/data/domain/user/corpus/spam and and nonspam
> directories. I run DSPAM with debugging on and so I see the "writing
> to corpus file ..." but if I understand it just writes the file and
> that's it. So again I don't get the point how these corpus files are
> useful to DSPAM? ;) Anyone can explain? I did not find any more infos
> in the README about that.
>

I'm not sure either, but google suggests:

http://osdir.com/ml/mail.spam.dspam.devel/2005-01/msg00021.html

So it helps you in building a corpus, it doesn't help DSPAM at all ;)

Tom

------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Dspam-user] makeCorpus=on

ML mail
Hi Tom,

Thanks for the hint. So I wasn't so wrong implying that the makeCorpus is useless for DSPAM itself.

Regarding TuM, I read more exactly and it mentions:

"It is a
          balance of resources as well, as only less-than-mature tokens are
          written to the database. NOTE: You should corpus train before
          using tum."

So I thought that I would need the makeCorpus for that but it looks like the README mentions another kind of corpus training unrelated to the makeCorpus parameter. Anyone has an idea what is meant here with "corpus train"?

Regards
ML


On Friday, December 13, 2013 12:42 PM, Tom Hendrikx <[hidden email]> wrote:
On 12/13/2013 11:34 AM, ML mail wrote:

> Hi,
>
> I just switched from TOE training mode to TUM in the hope of better
> results for new e-mail accounts. As recommended in the README I have
> turned on the makeCorpus preference.
>
>
> Now regarding to that makeCorpus preference I was wondering what is
> the point exactly in this option? or how is it useful to DSPAM? The
> only thing I notice DSPAM doing is saving the mails in
> /var/spool/dspam/data/domain/user/corpus/spam and and nonspam
> directories. I run DSPAM with debugging on and so I see the "writing
> to corpus file ..." but if I understand it just writes the file and
> that's it. So again I don't get the point how these corpus files are
> useful to DSPAM? ;) Anyone can explain? I did not find any more infos
> in the README about that.
>

I'm not sure either, but google suggests:

http://osdir.com/ml/mail.spam.dspam.devel/2005-01/msg00021.html

So it helps you in building a corpus, it doesn't help DSPAM at all ;)

Tom


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user



------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Dspam-user] makeCorpus=on

Dov Zamir
RE: [Dspam-user] makeCorpus=on

You can use the corpus you just built to pre-train another user, assuming most users have more or less the same type of spam and non-spam. There are curpuses (corpi?) out there but they are mostly too old to be of any use.

So, you can use dpam to create a corpus from a user you are training and then use this corpus to pre-train new users when the corpus is big enough.

 

-----הודעה מקורית-----
מאת: ML mail <[hidden email]>
נשלח: שיש 13-12-2013 13:57
נושא: Re: [Dspam-user] makeCorpus=on
אל: [hidden email];
Hi Tom,
 
Thanks for the hint. So I wasn't so wrong implying that the makeCorpus is useless for DSPAM itself.
 
Regarding TuM, I read more exactly and it mentions:
 
"It is a
          balance of resources as well, as only less-than-mature tokens are
          written to the database. NOTE: You should corpus train before
          using tum."
 
So I thought that I would need the makeCorpus for that but it looks like the README mentions another kind of corpus training unrelated to the makeCorpus parameter. Anyone has an idea what is meant here with "corpus train"?
 
Regards
ML


On Friday, December 13, 2013 12:42 PM, Tom Hendrikx <[hidden email]> wrote:
On 12/13/2013 11:34 AM, ML mail wrote:

> Hi,
>
> I just switched from TOE training mode to TUM in the hope of better
> results for new e-mail accounts. As recommended in the README I have
> turned on the makeCorpus preference.
>
>
> Now regarding to that makeCorpus preference I was wondering what is
> the point exactly in this option? or how is it useful to DSPAM? The
> only thing I notice DSPAM doing is saving the mails in
> /var/spool/dspam/data/domain/user/corpus/spam and and nonspam
> directories. I run DSPAM with debugging on and so I see the "writing
> to corpus file ..." but if I understand it just writes the file and
> that's it. So again I don't get the point how these corpus files are
> useful to DSPAM? ;) Anyone can explain? I did not find any more infos
> in the README about that.
>

I'm not sure either, but google suggests:

http://osdir.com/ml/mail.spam.dspam.devel/2005-01/msg00021.html

So it helps you in building a corpus, it doesn't help DSPAM at all ;)

Tom


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user

 
!DSPAM:1000,52aaf64469081702922955!
------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk

!DSPAM:1000,52aaf64469081702922955!
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user


!DSPAM:1000,52aaf64469081702922955!




************************************************************************************
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals & computer viruses.
************************************************************************************


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Dspam-user] makeCorpus=on

Tom Hendrikx
In reply to this post by ML mail
On 12/13/2013 12:50 PM, ML mail wrote:

> Hi Tom,
>
> Thanks for the hint. So I wasn't so wrong implying that the
> makeCorpus is useless for DSPAM itself.
>
>
> Regarding TuM, I read more exactly and it mentions:
>
> "It is a balance of resources as well, as only less-than-mature
> tokens are written to the database. NOTE: You should corpus train
> before using tum."
>
> So I thought that I would need the makeCorpus for that but it looks
> like the README mentions another kind of corpus training unrelated to
> the makeCorpus parameter. Anyone has an idea what is meant here with
> "corpus train"?

Corpus training means that need to have ham and spam corpi (which you
can build using an already running DSPAM instance, as we just found out
:>). With the corpi, you can train a new DSPAM install before it starts
processing mail. This is something you can do using dspam_train.

>
> Regards ML
>
>
>
>
> On Friday, December 13, 2013 12:42 PM, Tom Hendrikx
> <[hidden email]> wrote:
>
> On 12/13/2013 11:34 AM, ML mail wrote:
>> Hi,
>>
>> I just switched from TOE training mode to TUM in the hope of
>> better results for new e-mail accounts. As recommended in the
>> README I have turned on the makeCorpus preference.
>>
>>
>> Now regarding to that makeCorpus preference I was wondering what
>> is the point exactly in this option? or how is it useful to DSPAM?
>> The only thing I notice DSPAM doing is saving the mails in
>> /var/spool/dspam/data/domain/user/corpus/spam and and nonspam
>> directories. I run DSPAM with debugging on and so I see the
>> "writing to corpus file ..." but if I understand it just writes the
>> file and that's it. So again I don't get the point how these corpus
>> files are useful to DSPAM? ;) Anyone can explain? I did not find
>> any more infos in the README about that.
>>
>
> I'm not sure either, but google suggests:
>
> http://osdir.com/ml/mail.spam.dspam.devel/2005-01/msg00021.html
>
> So it helps you in building a corpus, it doesn't help DSPAM at all
> ;)
>
> Tom
>
>
> ------------------------------------------------------------------------------
>
>
Rapidly troubleshoot problems before they affect your business. Most IT
> organizations don't have a clear picture of how application
> performance affects their revenue. With AppDynamics, you get 100%
> visibility into your Java,.NET, & PHP application. Start your 15-day
> FREE TRIAL of AppDynamics Pro!
> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
>
>
_______________________________________________
> Dspam-user mailing list [hidden email]
> https://lists.sourceforge.net/lists/listinfo/dspam-user
>
>
>
> ------------------------------------------------------------------------------
>
>
Rapidly troubleshoot problems before they affect your business. Most IT

> organizations don't have a clear picture of how application
> performance affects their revenue. With AppDynamics, you get 100%
> visibility into your Java,.NET, & PHP application. Start your 15-day
> FREE TRIAL of AppDynamics Pro!
> http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
>
>
>
>
> _______________________________________________ Dspam-user mailing
> list [hidden email]
> https://lists.sourceforge.net/lists/listinfo/dspam-user
>


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Dspam-user] makeCorpus=on

ML mail
In reply to this post by Dov Zamir
Thanks I get it now for what the corpus are used. So to resume it doesn't help DSPAM itself unless you use dspam_train to train a user or a global group using that user generated corpus.



On Friday, December 13, 2013 2:08 PM, Dov Zamir <[hidden email]> wrote:
RE: [Dspam-user] makeCorpus=on
You can use the corpus you just built to pre-train another user, assuming most users have more or less the same type of spam and non-spam. There are curpuses (corpi?) out there but they are mostly too old to be of any use.
So, you can use dpam to create a corpus from a user you are training and then use this corpus to pre-train new users when the corpus is big enough.
 
-----הודעה מקורית-----
מאת: ML mail <[hidden email]>
נשלח: שיש 13-12-2013 13:57
נושא: Re: [Dspam-user] makeCorpus=on
אל: [hidden email];
Hi Tom,
 
Thanks for the hint. So I wasn't so wrong implying that the makeCorpus is useless for DSPAM itself.
 
Regarding TuM, I read more exactly and it mentions:
 
"It is a
          balance of resources as well, as only less-than-mature tokens are
          written to the database. NOTE: You should corpus train before
          using tum."
 
So I thought that I would need the makeCorpus for that but it looks like the README mentions another kind of corpus training unrelated to the makeCorpus parameter. Anyone has an idea what is meant here with "corpus train"?
 
Regards
ML


On Friday, December 13, 2013 12:42 PM, Tom Hendrikx <[hidden email]> wrote:
On 12/13/2013 11:34 AM, ML mail wrote:

> Hi,
>
> I just switched from TOE training mode to TUM in the hope of better
> results for new e-mail accounts. As recommended in the README I have
> turned on the makeCorpus preference.
>
>
> Now regarding to that makeCorpus preference I was wondering what is
> the point exactly in this option? or how is it useful to DSPAM? The
> only thing I notice DSPAM doing is saving the mails in
> /var/spool/dspam/data/domain/user/corpus/spam and and nonspam
> directories. I run DSPAM with debugging on and so I see the "writing
> to corpus file ..." but if I understand it just writes the file and
> that's it. So again I don't get the point how these corpus files are
> useful to DSPAM? ;) Anyone can explain? I did not find any more infos
> in the README about that.
>

I'm not sure either, but google suggests:

http://osdir.com/ml/mail.spam.dspam.devel/2005-01/msg00021.html

So it helps you in building a corpus, it doesn't help DSPAM at all ;)

Tom


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user

 
!DSPAM:1000,52aaf64469081702922955!
------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk

!DSPAM:1000,52aaf64469081702922955!
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user


!DSPAM:1000,52aaf64469081702922955!




************************************************************************************
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals & computer viruses.
************************************************************************************


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk

_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user



------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT
organizations don't have a clear picture of how application performance
affects their revenue. With AppDynamics, you get 100% visibility into your
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________
Dspam-user mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/dspam-user
Loading...