From: Thomas E. <Tho...@th...> - 2014-08-11 06:12:41
|
what is your setting for 'autoCorrectCorpus' ? - set it to the default value check the settings (..... down to MaxBytes) in the 'Collecting' section check the settings in the 'Rebuild SpamDB' section - stop assp - delete the file 'normfile' in the assp folder - start assp - force a rebuildspamdb - wait until it has finished - ignore the result - run a rebuildspamdb again Now the 'Corpus norm' and 'Corpus confidence' should be ~1 Thomas Von: Rob Ellsworth <ell...@gm...> An: For Users of ASSP <ass...@li...> Datum: 07.08.2014 15:10 Betreff: [Assp-user] ASSP v2 - Corpus norm WAY out of wack Hello everyone, We've been using ASSP v2 (2.4.1.14167) for a little over a month now. I rebuild my spam DB daily at midnight, and slowly but surely our Corpus Norm has climbed from 1.0 to an alarming 2.85. Here's the snippet from my rebuild logs: Aug-07-14 00:45:56 Spam Weight: 19,406,146 Aug-07-14 00:45:56 Not-Spam Weight: 6,809,812 Aug-07-14 00:45:56 Corpus norm: 2.8497 - (warning: extremely spam heavy) Aug-07-14 00:45:56 Corpus confidence: 0.01516294 Aug-07-14 00:45:56 Recommendation: RebuildSpamDB will limit the number of used messages in your corpus. Excess files will be ingored. Aug-07-14 00:45:56 Corpus norm should be between 0.6 and 1.4 Any ideas on what I can do to get the norm closer to an acceptable range? We do get a LOT of spam each day so this # doesn't surprise me. Thank you for taking the time to read this, - Rob ------------------------------------------------------------------------------ Infragistics Professional Build stunning WinForms apps today! Reboot your WinForms applications with our WinForms controls. Build a bridge from your legacy apps to the future. http://pubads.g.doubleclick.net/gampad/clk?id=153845071&iu=/4140/ostg.clktrk _______________________________________________ Assp-user mailing list Ass...@li... https://lists.sourceforge.net/lists/listinfo/assp-user DISCLAIMER: ******************************************************* This email and any files transmitted with it may be confidential, legally privileged and protected in law and are intended solely for the use of the individual to whom it is addressed. This email was multiple times scanned for viruses. There should be no known virus in this email! ******************************************************* |