This repo contains code for EMNLP 2021 paper: Uncovering Implicit Gender Bias Through Commonsense Inference

Getting Started

To get a local copy up and running follow these simple steps.

Prerequisites

This is an example of how to list things you need to use the software and how to install them.

• Prerequisites

  pip install -r requirements.txt

Download RocStory dataset from https://cs.rochester.edu/nlp/rocstories/

Download StanfordNERTagger

Installation

1. COMeT2

   Install COMeT2 according to https://github.com/vered1986/comet-commonsense

Usage

1. Classify stories according to protagonsit's gender

   python preprocess.py <story_filename.tsv>

2. Anonymization

   python replaceGender.py 

3. Extract stories having more than two characters

   python extractTwo.py 

4. Classify sentences according to protagonist

   python findSubj.py 

5. Get COMeT outputs

   python generate_inferences.py

6. Calculate Valence, arousal scores

   python connotation_COMET_NRC.py

7. Calculate Intellect, Appearance, Power scores

   python get_lexicon_average.py

   Acknowledgement:

   We borrowed some code from this repository: https://github.com/ddemszky/textbook-analysis