Re: Mixed 64-bit system GerH binaries / BillYscripts

Paolo escreveu:
> On Sat, Aug 30, 2008 at 08:27:22PM -0400, Bill Yerazunis wrote:
[...]
> 
>> But then OSB came up (thanks Christian!) and that cut the size of
> 
> That was Fidelis ;)

Thanks for the correction, Paolo. That's a recurring misinformation.

I'm proud I've contributed the OSB idea and patches to what is now the
default classifier in CRM114. Below is a quote from "jgc's spam and
anti-spam newsletter #55" (01 May 2007), which I hope makes it clear
once and for all :-)

---------------------------------------------------------------------
Last time I mentioned the DSPAM 3.8.0 update using the DSPAM author's
own words:

  DSPAM 3.8.0 is out with major feature enhancements: This is an
  unofficial release, but significant. Several significant bugfixes
  have been made. Bill Yerazunis' Orthogonal Sparse Bigram (OSB)
  tokenizer algorithm has been added. The code has been significantly
  optimized to run much faster and with fewer resources.

Fidelis Assis writes:

  I think there's a confusion here, which tokenizer is DSPAM going to
  use, Bill's one or OSB? Bill's tokenizer is called SBPH; OSB is
  mine. A comparison between both techniques can be found in the
  articles below:

    http://www.siefkes.net/ie/winnow-spam.pdf
    http://osbf-lua.luaforge.net/papers/OSBF-Lua_VBFeb07.pdf

And Christian Siefkes adds:

  just a quick correction: OSB has been invented by Fidelis Assis, not
  by Bill (though it was first employed in CRM114 due to patches
  contributed by Fidelis).
---------------------------------------------------------------------

BTW, those on this list using OSB and OSBF might like to know that "CEAS
2008 Spam Filter Live Challenge" results have just been published and
that OSBF-Lua (TREC06 version) was the #1 by both criteria: LAM% and 1-AUC%.

  http://www.ceas-challenge.cc/results/
  http://www.ceas-challenge.cc/

OBS: TREC06 OSBF-Lua, like TREC07 CRM114 and some others on that page,
were entered by the organizers for comparison purposes and considered
unofficial competitors (reference).

-- 
Fidelis Assis


-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/