[Spambayes] Duplicate messages treated differently...?
Kenneth Sole
sole at soleassociates.com
Sat Feb 12 23:34:49 CET 2005
Hello Tony and all,
I had posted earlier about this issue, but then it did not recur...
Until today, that is:
On occasion, I receive duplicate Spams. They are identical in every
detail, but one is moved to my Spam folder, and the other is moved to
my Review folder.
Each of the messages has the same Combined Score and clues summary:
Combined Score: 45% (0.447498)
Internal ham score (*H*): 0.371119
Internal spam score (*S*): 0.266116
# ham trained on: 143
# spam trained on: 713
42 Significant Tokens
token spamprob #ham #spam
'floured' 0.0652174 3 0
'bi:and let' 0.14447 2 1
'process.' 0.149485 3 2
'contributes' 0.155172 1 0
'letting' 0.155172 1 0
'skip:o 10' 0.173445 33 34
'ease' 0.200725 2 2
'bi:love the' 0.228209 1 1
'bi:url:r url:html' 0.228209 1 1
'when' 0.228397 45 66
'skip:a 20' 0.229858 7 10
'skip:k 10' 0.269238 4 7
'give' 0.287877 20 40
'skip:n 10' 0.301553 15 32
'skip:u 10' 0.30564 22 48
"you'll" 0.317892 7 16
'skip:m 10' 0.343713 20 52
'skip:s 10' 0.358915 23 64
'skip:e 10' 0.361182 21 59
'bi:the best' 0.36651 7 20
'skip:t 10' 0.376812 12 36
'care' 0.378393 5 15
'skip:p 10' 0.379167 26 79
'skip:c 10' 0.379495 44 134
'bi:url:html proto:http' 0.385615 9 28
'skip:i 10' 0.386699 22 69
'designer' 0.388231 1 3
'stop' 0.628758 4 34
'bi:to:2**0 x-mailer:none' 0.646495 40 365
'bi:header:Received:2 header:From:1' 0.720726 11 142
'subjectcharset:iso-8859-1' 0.7857 1 19
'bi:skip:s 10 skip:r 10' 0.844828 0 1
'bi:subject: subject:@' 0.844828 0 1
'cordage' 0.844828 0 1
'subject:&' 0.844828 0 1
'subject:Drug' 0.844828 0 1
'subject:hours' 0.844828 0 1
'subject:ordering' 0.844828 0 1
'sutra' 0.844828 0 1
'toe' 0.844828 0 1
'bi:skip:c 10 skip:i 10' 0.908163 0 2
'bi:virus:src="cid: proto:http' 0.990798 0 24
Message Stream
Return-Path: <saundramiltoniv at airagency.ru>
Delivered-To: soleassociates.com-sole at soleassociates.com
Received: (qmail 29838 invoked by uid 107); 12 Feb 2005 22:07:49 -0000
Received: from abayonne-152-1-73-186.w83-200.abo.wanadoo.fr (HELO
reset.ru)
(83.200.251.186)
by s9006.hostcentric.net with SMTP; 12 Feb 2005 22:07:49 -0000
Subject:
=?iso-8859-1?B?UHJlc2NyaXAmdGlvbiBEcnVnQHMgd2l0aGluIDI0IGhvdXJzIG9mIG9
yZGVyaW5nISAgICAgZA==?=
To: sole at soleassociates.com
From: "Saundra Milton" <saundramiltoniv at airagency.ru>
Message-ID: <216301c511b5$3c4bfd4a$36483149 at bagcp4g>
MIME-Version: 1.0
Date: Sun, 13 Feb 2005 10:16:01 +0000
Content-Type: multipart/related; type="multipart/alternative";
boundary="----=_NextPart_000_0E8C_B65A3F86.CF183D98"
You'll love the ease of our ordering process. Give us a chance, and
let the best take care of you!
azyme photonephograph refractory coarsening extractors oxalis
indesirable bribetaker. scuta apaches roughtailed androgens letting
commissionship proddle colorationally ignatia. optative brimfully
compunctionary floured. lenities semimalignant cellarets quillaia
prochein accumulativ eucre. hooping flacons legitim undisadvantageous
ingenerability optative extrapelvic nonputrescent cathedralic khis.
plafond pullers pondweed sharpness prestomium urediniosporic
ultramodernist purrone spallation autocholecystectomy.
incapabilities phoenicians devilish officeholders. javeline
antiquarians blackguard eupneic oomiak pinks alulet toe coalescency
unwiseness. slopeness cellarets. tetramethylammonium prealarm brevit
salinizing mirv individualize bougar regulates pavanes. undespondently
micromineralogical scrawler imperforate streptobacillus unglorifying
preilluminate hexahydrated enkindling legitim. cicutoxin ciboule
typhuses complexionless iridioplatinum hiccupped fourcher. contributes
sulfobenzoic when glebous cordage incautiousness homolog remass eucre
dooryards. tannings appendix volplane washier audiotypist
iridioplatinum precanceling hilting omnirepresentative. sheiklike
subsulphid besprinkling implicational megalopsia enameler offlet
kinaesthetically alulet moonwalking interpolater. finched tonsures
cytost kinkle contraremonstrance prisonlike fistiana synclitic.
heirskip aversation cellarets trysts photokinesis figurize ravisher
hijacked. eyedropperful floured malocclusion tolerability eroses
demurely luncheoner. pokerish simianity slopeness unperpetrated.
foreprovision hysterical zimentwater underclothing samburs. sheiklike
fetid uterointestinal sutra eme designer thysanopteran indesirable
outpick. hysterical bimasty zoonomia naturality eliciting antischool
altarpieces vaugnerite overshooting. encapsulating socializable
battlemented oysterhood haha. scincoids superimposing reinstauration
ultraenthusiastic nonartistic.
Click muzjiks taurocholate me to stop this...
<!doctype html public "-//w3c//dtd html 3.2//en">
<html><head>
<META content="MSHTML 6.00.2800.1226" name=GENERATOR></head>
You'll love the ease of our ordering process. Give us a chance, and
let the best take care of you!<br>
<a href="http://CQAESJCC.pharmz1.com/p1/pk4/HVEVEFBGFWP.html"><img
src="cid:d92501c511b5$55a44be0$144dfdf7 at gk7vq53"></a>
<br><br><br><br>azyme photonephograph refractory coarsening extractors
oxalis indesirable bribetaker. scuta apaches roughtailed androgens
letting commissionship proddle colorationally ignatia. optative
brimfully compunctionary floured. lenities semimalignant cellarets
quillaia prochein accumulativ eucre. hooping flacons legitim
undisadvantageous ingenerability optative extrapelvic nonputrescent
cathedralic khis. plafond pullers pondweed sharpness prestomium
urediniosporic ultramodernist purrone spallation
autocholecystectomy.<br>incapabilities phoenicians devilish
officeholders. javeline antiquarians blackguard eupneic oomiak pinks
alulet toe coalescency unwiseness. slopeness cellarets.
tetramethylammonium prealarm brevit salinizing mirv individualize
bougar regulates pavanes. undespondently micromineralogical scrawler
imperforate streptobacillus unglorifying preilluminate hexahydrated
enkindling legitim. cicutoxin ciboule typhuses complexionless
iridioplatinum hiccupped fourcher. contributes sulfobenzoic when
glebous cordage incautiousness homolog remass eucre dooryards.
tannings appendix volplane washier audiotypist iridioplatinum
precanceling hilting omnirepresentative. sheiklike subsulphid
besprinkling implicational megalopsia enameler offlet kinaesthetically
alulet moonwalking interpolater. finched tonsures cytost kinkle
contraremonstrance prisonlike fistiana synclitic. heirskip aversation
cellarets trysts photokinesis figurize ravisher hijacked.
eyedropperful floured malocclusion tolerability eroses demurely
luncheoner. pokerish simianity slopeness unperpetrated. foreprovision
hysterical zimentwater underclothing samburs. sheiklike fetid
uterointestinal sutra eme designer thysanopteran indesirable outpick.
hysterical bimasty zoonomia naturality eliciting antischool
altarpieces vaugnerite overshooting. encapsulating socializable
battlemented oysterhood haha. scincoids superimposing reinstauration
ultraenthusiastic nonartistic.<br><br>
<br><br>
Click <a href="http://CFAKRDCIR.pharmz1.com/r.html">muzjiks
taurocholate me</a> to stop this...
</html>
and here is the most recent log:
Loaded bayes database from 'C:\Documents and Settings\Kenneth
Sole\Application Data\SpamBayes\default_bayes_database.db'
Loaded message database from 'C:\Documents and Settings\Kenneth
Sole\Application Data\SpamBayes\default_message_database.db'
Bayes database initialized with 712 spam and 142 good messages
SpamBayes Outlook Addin Binary Version 1.0.2 (January 2005) starting
(with engine SpamBayes Engine Version 0.3 (January 2004))
on Windows 5.0.2195 (Service Pack 4)
using Python 2.4 (#60, Nov 30 2004, 11:49:19) [MSC v.1310 32 bit
(Intel)]
SpamBayes: Watching (for filtering) in 'Desktop/Inbox'
SpamBayes: Watching (for incremental training) in 'Desktop/Junk SPAM'
Processing missed spam in folder 'Inbox' by starting a timer
Message 'fwd: my new contact info' in 'Desktop/Inbox' had a Spam
classification of 'Yes'
Message 'check the site for great offfer on meds' in 'Desktop/Inbox'
had a Spam classification of 'Yes'
Message 'F-R-E-E Adult personals' in 'Desktop/Inbox' had a Spam
classification of 'Unsure'
FAILED to add the toolbar item 'SpamBayesCommand.Manager' -
(-2147352567, 'Exception occurred.', (0, None, 'Member not found.',
None, 0, -2147467259), None)
The above toolbar message is common - recreating the toolbar...
Message 'dude, i wrote you 10 times...' in 'Desktop/Inbox' had a Spam
classification of 'Unsure'
Training on message 'dude, i wrote you 10 times...' in 'Desktop/Junk
SPAM - trained as spam
Training on message 'Symantec Gold Enterprise Support Phone Numbers'
in 'Desktop/Inbox - trained as good
Message 'Prescrip&tion Drug at s within 24 hours of ordering! d' in
'Desktop/Inbox' had a Spam classification of 'Unsure'
Please note that the last line above refers to the message(s) in
question, but seems to indicate that it arrived only once and was
classified as "unsure."
I am certain that it is the same message by the way. The two that
appear have the same I.D. in the header.
Of course, after looking at the Clues, it would appear that the
message belongs in my Review folder only.
And finally, I run only two extremely simple rules. They each look for
certain words in the subject line and move messages meeting that
criterion to particular folders (other than my review, or spam
folders.)
As I mentioned earlier, this is no disaster, but it is odd, and I
would be curious to know what might be causing this problem.
Sincere thanks for any help on this,
--
Kenneth
Sole & Associates, Inc.
Box 292
Durham, New Hampshire 03824
Voice: 603-659-3169
Fax: 603-659-2248
Email: sole at soleAssociates.com
URL: http://www.soleAssociates.com
PGP:
http://wwwkeys.ch.pgp.net:11371/pks/lookup?op=get&search=0xE17941C6
More information about the Spambayes
mailing list