I did some preliminary analyses on our data using NLTK, a Natural Language Processing package for Python. Mostly analyzing phrases that repeat themselves.
I put them online in a couple of IPython Notebooks.
First one is mostly just me messing around with NLTK functions on a small subset of the data (partially to give a tour of NLTK to folks who haven’t played with it):
This one goes deeper into the data, analyzing longer phrases and the whole body of entries: