spam corpus
Normal, when you get spam email, you will delete that, or mark that as spam, so the program (e.g., Fox) will do machine learning and remember that. I think, in the world, a lot of person are attacked by spam email. but in the other hand, a lot of person are sending spam email, maybe. in different country with different language. If you collect the spam email and make that as corpus, and then translate (of course machine translation) to other language, it would get a lot of users or more practical.
Along with the internationalization of spam email , the machine translation will get developed rapidly, especially, AV machine translation, or AV robot.
It should be fun.
Comment?