Re: Problem with BayesIT 0.4gm
Hello Alexey, * on Friday (9th January 2004 - 21:38) Alexey N. Vinogradov wrote: ANV What is the size of your spam/non-spam bases according to statistic? And also, ANV what is the size of base files? I find a very seldom intention that on some ANV conditions BayesIt can collapse it's own base, but I can't yet reproduce this... I don't have the details any more because I now use K9 again, but the databases must have been rather big. My spam mail folder now has over 8.000 mails in it an the good mails aterr more the 10 times this amount. -- Best Regards | OS : Windows XP 5.1.2600.Service Pack 1 | MUA: TB! 2.02.3 CE - S/N: 058F9794 Matthias Frost | 10.01.2004 at 14:31 (GMT +0100) == * Virus-Checked by AntiVir - using newest program database updates * Current beta is 2.03 Beta/28 | Using TBBETA information: http://www.silverstones.com/thebat/TBUDLInfo.html
Re: Problem with BayesIT 0.4gm
Hello Martin, * on Friday (9th January 2004 - 14:04) Martin Webster wrote: MW How many spam messages do you receive daily? appox. 150 per day. 1/3 of those are spam mails. MW One possible reason is you're receiving a new breed of spam mail, which MW has an HTML spam message plus a text part with a random list of words; MW e.g. addition congruent malnourished masterpiece patterson planetaria MW legacy washboard coerce prognosticate rose and so on. MW Maybe your Bayesian filter needs a little more training? As sais my spam dtabase now has over 8.000 mails in it. If at all then I think it is a bit over-trained :) -- Best Regards | OS : Windows XP 5.1.2600.Service Pack 1 | MUA: TB! 2.02.3 CE - S/N: 058F9794 Matthias Frost | 10.01.2004 at 14:33 (GMT +0100) == * Virus-Checked by AntiVir - using newest program database updates * Current beta is 2.03 Beta/28 | Using TBBETA information: http://www.silverstones.com/thebat/TBUDLInfo.html
Re: Problem with BayesIT 0.4gm
Hello MikeD, * on Friday (9th January 2004 - 10:28) MikeD (3) wrote: M3 It sounds like perhaps the database that Beysit creates has gotten M3 corrupt. Did you keep the training base? If so, just re-run the M3 'create' function to regenerate it. If not, you will need to save up M3 some more spam and then retrain it. I did so but is had not helped. I now use K9 again because I get over 50 spam mails per day. -- Best Regards | OS : Windows XP 5.1.2600.Service Pack 1 | MUA: TB! 2.02.3 CE - S/N: 058F9794 Matthias Frost | 10.01.2004 at 14:34 (GMT +0100) == * Virus-Checked by AntiVir - using newest program database updates * Current beta is 2.03 Beta/28 | Using TBBETA information: http://www.silverstones.com/thebat/TBUDLInfo.html
RTF Viewer Signatures
Good evening Beta-Testers, has anybody also noticed an odd behaviour of the rich text viewer concerning signature formatting? For example if you setup a different text colour for signatures and there is more then one -- in a displayed mail then only the text below the last -- is displayed in your chosen colour. -- Best regards Matthias Frost __ 14 Dec 2001 18:37 TB! V1.54 Beta/15 [85790CB6] -- _ Archives : http://tbbeta.thebat.dutaint.com Moderators : mailto:[EMAIL PROTECTED] Unsubscribe: mailto:[EMAIL PROTECTED] Latest Beta: 1.54 beta/15 Wish List : http://wish.thebat.dutaint.com
