Hello ricardo,
Monday, January 26, 2004, 7:00:03 PM, you wrote:
r I'm attaching a message (I hope it makes it through to the list) which
r scored a low 0.1 with SA 2.63.
r Does anyone have any suggestions on how to possibly make SA get a higher
r score for this type of message? Any new recipes
Hi,
I'm attaching a message (I hope it makes it through to the list) which
scored a low 0.1 with SA 2.63.
Does anyone have any suggestions on how to possibly make SA get a higher
score for this type of message? Any new recipes that might improve the
scoring?
Thanks
RicardoDelivered-To:
Ok, interesting.
But I do have bayes trained with many thousands of messages;
I imagine many of them having similar stuff, and it seems
that the bayes score for this message was quite low.
How does the Bayes training work, anyway... If this one
message gets trained as --spam, how much of an
At 08:25 AM 1/27/04 -0800, Ricardo Kleemann wrote:
How does the Bayes training work, anyway..
In short:
First, you need to understand bayes is based on breaking email down into
tokens. For simplicity, you can just consider each word of an email to be
a token. SA uses other tokens (header
At 07:00 PM 1/26/04 -0800, ricardo wrote:
Does anyone have any suggestions on how to possibly make SA get a higher
score for this type of message? Any new recipes that might improve the
scoring?
Quite frankly, that email with all its mis-spelled words should be easy
pickings for bayes. Train.
