Spam collection

[Mikkel Rasmussen]

Disclaimer: This is not just for Python programmers (since I use Python I
thought it would be nice to co-operate with other Python programmers).

I have thought about sharing my spam collection with others for use in
developing a better spam filter. We need a large collection of spam to be
able to do various forms of analysis on it. I don't know if such a
collection already exists. If so, I would like to add mine.

My spam filter "idea" is to use keywords, because I use Outlook and Outlook
does not give any other possibilities (as far as I know). The problem is in
choosing the best keywords without using *any* word that occurs in a
non-spam message.

We probably also need a definition of spam. A tentative definition could be
"irrelevant messages" where irrelevant gives a subjective perspective. My
spam might not be your spam :-)

Any further ideas?

Mikkel Rasmussen