Ended up using a <
instead of a less than sign.
So then the tag showed up.
So at least I know that part is working.
Im looking at the emails that come in and
I cant tell any difference between the spam
and the ham. Each has the same balance of
numeric values. Some are 0.0 and spam
and some are 0
WoW,
Interesting. When I try to uses
the less than sign as u describe, I get exceptions
spitting out. Cant process config.xml..
David Legg wrote:
>
> Joe,
>
>> Viewing the config.xml in firfox works OK.
>>
>
> That's a good test.
>
>> SO I did a little test. OI attempted to add a head
Joe,
Viewing the config.xml in firfox works OK.
That's a good test.
SO I did a little test. OI attempted to add a header where I conjured up
a name and set the
X-MessageIsSpamProbability > 0.05" to add the header.
Still nothing in the message..
As a quick check why not tempora
Viewing the config.xml in firfox works OK.
I do all my edits using the vi editor and make sure no
extra chars are on the end of line. SO I dont think thats it, unless a
line that is over run by real long xml string is broken and still viewable
from firefox??? I have seen that before. Using you
Kent,
and if these exist during the Spam training, i.e. 100% of your example
Spam emails contain these tokens, perhaps this is tainting the token
ratings?
Idea + wild guess = HTH
Thanks for the ideas. Let's hope it is simply down to a mis-configured
config.xml. Although, even if it is i
Joe,
I admire your tenacity ;-)
Making progress!
I pulled in the perts from your config.xml that looked like thety applied
to bayesian analysis. Now Im seeing numeric values. However,
the X-MessageIsSpam tag is missing from ALL messages.
I do see all the values in the
X-MessageIsSpamProbabili
Making progress!
I pulled in the perts from your config.xml that looked like thety applied
to bayesian analysis. Now Im seeing numeric values. However,
the X-MessageIsSpam tag is missing from ALL messages.
I do see all the values in the
X-MessageIsSpamProbability: 1.0826917212753594E-212
tag, in
Making progress!
I pulled in the perts from your config.xml that looked like thety applied
to bayesian analysis. Now Im seeing numeric values. However,
the X-MessageIsSpam tag is missing from ALL messages.
I do see all the values in the
X-MessageIsSpamProbability: 1.0826917212753594E-212
tag, in
tainting the token
ratings?
Idea + wild guess = HTH
Kent
-Original Message-
From: David Legg [mailto:david.l...@searchevent.co.uk]
Sent: Tuesday, February 10, 2009 1:57 PM
To: James Users List
Subject: Re: How do I reduce SPAM
> Here is what I see as well; ( o
Joe,
If it helps I have posted a summary of my working config.xml settings on
this list before. Here they are again. I've removed all the
non-Bayesian bits and replaced them with a line of dots. Maybe it will
give you some ideas.
config.xml ---
Joe,
I think I may have it!WHich is it?
I found two different versions of the compare;
I spotted that too but didn't mention it because I didn't think it would
make any difference.
In my working config I use: -
onMailetException="ignore">
db://maildb
I think I may have it!WHich is it?
I found two different versions of the compare;
1. mine
--
true
X-MessageIsSpam
true
2. Official james site
---
ALL messages contain these tags. The emails are a mixture. Some are
plain test and some html. There are some that come through with
pictures,attachments,
and the stnadard hoopla. I must have something wrong in my config.xml.
Why would the NaN always show up.
David Legg wrote:
>
>
>> Her
Here is what I see as well; ( on ALL messages)
X-MessageIsSpamProbability: NaN
X-MessageIsSpam: true
Mmm... Ok. Well, as you may know 'NaN' is short for 'Not a Number' in
floating point speak. So something has caused the spam probability
value to be such a large or small number that J
Yes that is there too;
I left it off,
I dont run spamassasin that I know of. Could be in front of my serves,
other landing zones..
Here is what I see as well; ( on ALL messages)
X-MessageIsSpamProbability: NaN
X-MessageIsSpam: true
David Legg wrote:
>
> Joe,
>
>> Here is what is in the em
Joe,
Here is what is in the emails that are getting tagged bad;
X-Spam-Score: -2.6
X-Spam-Report: -2.6 BAYES_00 BODY: Bayesian spam probability
is 0 to 1%
[score: 0.]
If you can't see any header in the email called
'X-MessageIsSpamProbability' then you aren't loo
Here is what is in the emails that are getting tagged bad;
X-Spam-Score: -2.6
X-Spam-Report: -2.6 BAYES_00 BODY: Bayesian spam probability
is 0 to 1%
[score: 0.]
jbi joe wrote:
>
> IM getting way too much spam. I have tried to train the bayesian analysis
> mailet by
Joe,
jbi joe wrote:
UPDATE
Thought I would provide an update to this..
WHAT could I be doing wrong.
It sounds very strange.
I think you need to check on the X-MessageIsSpamProbability headers of a
few emails in your deadletter repository to see what level of
probability the Bayesian filt
UPDATE
Thought I would provide an update to this..
Feeding spam to my "spam" email causes all email to be thought of
as spam wether it is or isnt. SO I started feeding more "HAM" in thinking
that maybe there need to be more training for HAM. Still ALL email
coming in is thougt to be spam and dum
Hi Joe,
Looks like feast or famine! I think I have it working.
Great! I knew you could do it.
I can see entries in the bayesiananalysis_spam table when I send in a
spam record.
Good, that's proof enough that your spam attachments are being processed
by the feeder.
However, one su
Looks like feast or famine! I think I have it working.
I can see entries in the bayesiananalysis_spam table when I send in a
spam record.However, one such spam causees all subsequent
emails to be considered spam. It looks like some kinda spam prevenetion
detractor. It circumvents everythi
Hi Joe,
Digging in DEEPER yields the following;
1. mailet log says there is no mail server named hostname.myfirst.domain so
I change my config.xml to just s...@mydomain.com for bayesian analysis. (
What should I be forwarding my spam to???
DO I need to purchase another domain for this to wor
Just a few words for you.
Spamhaus
Spamcop
SPF (Sender Policy Framework)
Reverse Pointer Lookup
- Original Message -
From: "jbi joe"
To:
Sent: Saturday, January 31, 2009 4:40 AM
Subject: How do I reduce SPAM
IM getting way too much spam. I have tried to train the bayesian analy
Digging in DEEPER yields the following;
1. mailet log says there is no mail server named hostname.myfirst.domain so
I change my config.xml to just s...@mydomain.com for bayesian analysis. (
What should I be forwarding my spam to???
DO I need to purchase another domain for this to work?
2. Al
jbi joe wrote:
Corpus is empty.. I see 5 tables in db.
All of them are empty.
| bayesiananalysis_ham |
| bayesiananalysis_messagecounts |
| bayesiananalysis_spam |
| deadletter |
| spool
Not sure why..
How can I debug?
Are you sure you are forwa
Corpus is empty.. I see 5 tables in db.
All of them are empty.
| bayesiananalysis_ham |
| bayesiananalysis_messagecounts |
| bayesiananalysis_spam |
| deadletter |
| spool
Not sure why..
How can I debug?
David Legg wrote:
>
> Hi Joe,
>
> jbi joe wro
Hi Joe,
jbi joe wrote:
IM getting way too much spam. I have tried to train the bayesian analysis
mailet by feeding it spam and notspam. However, it does nothing.
I haven't looked too closely but I did spot that your config file has
the following setting: -
true
This setting will mean
27 matches
Mail list logo
