[4/19/21] DEPRECATED -- please use this repo instead
This repo is no longer under active development.
This repo contains code for the paper, "Deconfounded Lexicon Induction for Interpretable Social Science".
You can read the paper here. It outlines the algorithms implemented in this repo.
The project website is here.
pip install -r requirements.txt
python main.py --config sample_config.json [--train] [--test]
Logs, models, features, and summary files will be written in the config's working_dir
. See sample_config.json
for an example config and explanation.