Deep Q-Learning for Event Summarization

Repo currently has 3 folders

Code
Paper
Presentation

This is the repository for our {Francisco Javier Arceo, Chris Kedzie} implementation of our Deep Q-Network using a Long Short Term Memory (DQN-LSTM) for Event Summarization.

To run the code simply enter in the command line:

th Code/runModelProd.lua --model lstm

The relevant code files in the /Code folder are

1. Utils/build_data.py
2. utils.lua
3. utilsNN.lua
4. runModelProd.lua
5. runModel.lua

The files in the /Paper folder are related to the TeX document to create the paper.

License

MIT

Notebooks

1. Run FinalProcessing.ipynb
2. Pytorch DQN Summarization.ipynb