Dependencies:
Natural Language Toolkit + Wordnet data

*JSON export:
jstree library included in tree_js

*DB export:
sqlalchemy, extensions for desired db (eg. python-mysqldb)


Program files:
tree.py: an implementation of trees
hyponym_generator.py: a generic generator for both json and db based branches
js_hyponym_trees.py: inherits from tree.py and hyponym_generator.py and makes them export to json
db_hyponym_trees.py: inherits tree.py and hyponym_generator.py and makes them export to a database

Data files:
whitelist.csv: a file to contain existing categories so that they can be marked with '@' during tree generation
word-totals.txt: a file of unigrams occuring more than 40,000 times in the google unigram corpus

Set-up:

Install nltk and its dependencies with pip:
$pip install -U numpy
$pip install -U pyyaml
$pip install -U nltk

Download wordnet corpora:
$python
>>>import nltk
>>>nltk.download()
Download wordnet and wordnet_ic from the GUI interface

Instructions:
Run either js_hyponym_trees.py for json export or db_hyponym_trees.py for export to a db via sqlalchemy