Once you select what is spam and what's not, and you have two new mboxes, you'd better off doing a tour on them with sa-learn from the SpamAssassin distribution. You'll see that your spamassassin will have far less false positives

From the sa-learn man page:

NAME
       sa-learn - train SpamAssassin's Bayesian classifier

SYNOPSIS
       sa-learn [options] [file]...

[...]
       Options:

        --ham                             Learn messages as ham (non-s
+pam)
        --spam                            Learn messages as spam
[...]
        --mbox                            Input sources are in mbox fo
+rmat
        --showdots                        Show progress using dots
        --no-rebuild                      Skip building databases afte
+r scan
[...]

DESCRIPTION
       Given a typical selection of your incoming mail classified
       as spam or ham (non-spam), this tool will feed each mail
       to SpamAssassin, allowing it to 'learn' what signs are
       likely to mean spam, and which are likely to mean ham.

       Simply run this command once for each of your mail fold­
       ers, and it will ''learn'' from the mail therein.

[...]
       SpamAssassin remembers which mail messages it's learnt
       already, and will not re-learn those messages again,
       unless you use the --forget option.  Messages learnt as
       spam will have SpamAssassin markup removed, on the fly.
[download]

Way aheaf of you. After 'manually' split the SPAM from regular mail, I released sa-learn off course.
Don't want this to happen a second time.



--
My opinions may have changed,
but not the fact that I am right