Bayesian spam scores for individual words

ffalcon · Post by **ffalcon** » February 20th, 2004, 10:13 am

Is there any way to get an analysis of my training.dat to find out which words are flags for spam or ham, and with what probability?

Thanks.

ffalcon · Post by **ffalcon** » February 21st, 2004, 8:32 pm

Found a solution:
http://bayesjunktool.mozdev.org/

You need java installed and in the path. There is discussion of how to get the application working that you can find by searching google groups for "bayes junk tool", and it's the first hit.