I am working on the popular 20 Newsgroup dataset, which can be found in the following link: http://qwone.com/~jason/20Newsgroups/
I have split this data into training
