> I gave this a whirl. Not so good for the samples I pulled > out of my current spam database.
Considering the way SpamBayes works, I think these results are very acceptable. They're "unlikely" words, so will yield pretty good results after a few of these ad-images' text are trained. Stuff like "EN_OY", "_EXl", "blgge__", "_he", "vlbra_orl" will likely apear consistently for all spam images that are converted, and unlikely to appear in legitimate images. Regards, Shawn K. Hall http://12PointDesign.com/ http://ReliableAnswers.com/ _______________________________________________ [email protected] http://mail.python.org/mailman/listinfo/spambayes Check the FAQ before asking: http://spambayes.sf.net/faq.html
