C-Command Software Forum

Message marked as mild spam cannot be trained as good

This message shows up with with a yellow tint in my spam folder. When I ask to “Train as Good”, nothing happens. Here’s the log:

Predicted: Spam (96)
Subject: CC: Match.com Message: RE: Thanks for the wink
From: cycletogether@talkmatch.com
Identifier: 2PMBpHgi57TcrKUlaRVwuw==
Reason: P(spam)=1.000[1.000], bias=0.508, I:9ba(1.000), A:cycletogether(0.999), F:cycletogether@(0.999), F:cycletogether@(0.999), cyclist(0.999), cyclist(0.999), I:3ae(0.999), C:922-07f(0.999), x-emailid:2699ae(0.999), x-campaign-data:MA-12292011-5-DBNS-129-abfff(0.999), 110500(0.999), I:07f3(0.999), x-campaign-data:2-3ae(0.999), x-emailid:3-4018-aab(0.999), U:abfff922(0.999)
Date: 2011-12-29 17:09:55 +0000

When you train the message as good, do you see any error messages in the “All Messages” section of the Console application?

Where is that located?

I am now unable to train any message as good…

After you open the regular SpamSieve log you will be in the Console application. Look at the top left of the log list, and click on All Messages.

You may need to quit SpamSieve and your mail program and replace the SpamSieve application file with a freshly downloaded copy.

If I reinstall, will I lose my existing corpus? That took three months to gather.

No; you’ll keep all the training data and statistics.