Bug 3460 - use 'Improved Chi' technique in classifier
Summary: use 'Improved Chi' technique in classifier
Alias: None
Product: Spamassassin
Classification: Unclassified
Component: Learner (show other bugs)
Version: SVN Trunk (Latest Devel Version)
Hardware: Other other
: P5 normal
Target Milestone: Future
Assignee: SpamAssassin Developer Mailing List
Depends on:
Blocks: 4560
  Show dependency tree
Reported: 2004-06-02 12:20 UTC by Justin Mason
Modified: 2012-01-18 21:52 UTC (History)
0 users

Attachment Type Modified Status Actions Submitter/CLA Status
Article in PDF Format application/pdf None Kevin A. McGrail [HasCLA]

Note You need to log in before you can comment on or make changes to this bug.
Description Justin Mason 2004-06-02 12:20:55 UTC
_Handling Redundancy in Email Token Probabilities_, Gary Robinson: 


Bogofilter results:


looks good.  one for the 3.1.0 queue I think...
Comment 1 Daniel Quinlan 2004-08-27 17:18:41 UTC
more accuracy and performance bugs going to 3.1.0 milestone
Comment 2 Daniel Quinlan 2005-03-29 16:08:46 UTC
assigning to 3.2.0, no action taken so far, but seems useful
Comment 3 Justin Mason 2006-05-26 10:16:05 UTC
This was a suggested idea for the Google Summer of Code 2006;
I'm adding it to the bugzilla for future use, and in case anyone feels
like implementing it.

Subject ID: spamassassin-improved-chi
Keywords: bayes, classifiers, chi, statistics, perl
Description: Has shown good results, reportedly.  Implement in SpamAssassin and
benchmark results using 10-fold cross-validation.
Possible Mentors: Justin Mason (jm at jmason.org)
Comment 4 Justin Mason 2006-12-05 05:51:55 UTC
moving off to 3.3.0, I doubt it's going to happen for 3.2.0... feel free to
retarget, of course
Comment 5 Justin Mason 2006-12-05 06:10:52 UTC
really move this time
Comment 6 Justin Mason 2009-06-29 04:26:49 UTC
not going to happen in time
Comment 7 Kevin A. McGrail 2012-01-18 21:52:45 UTC
Created attachment 5037 [details]
Article in PDF Format