I analyzed a set of tweets using the tm package, I cleaned the set, removed its scarcities & etc. I ended up with 30 terms only, on a set of about 10k tweet
tm