It detects if a line of tweet text is medically relevant or not.
So first run "read.py" and it generates the data, then run "train.py".
pu.db
will open up after training is over.
"out" will give you the precision and recall and "score[1]" will give you the accuracy