pronounced "preh-si"

• Currently the feature weighing is achieved using WordNet, but this will eventually be replaced with word2vec
• Still a lot of performance enhancements required. More tests and more API features to follow

• Install Python 2.7.x
• Install virtualenv to manage packages
• Install packages from eggs.txt
• Install nltk data packages, from prompt execute these

pip install virtualenv
virtualenv .env # this creates environment for this application
source .env/bin/activate
pip install -r eggs.txt
python nltk_setup.py

or Use make shipped with this project

make env
source .env/bin/activate

• After the setup you can run the following script against a document directory with a list of documents belonging to similar topic to generate Summary

./summarize.py <content directory path>

make tests