Run in the follwing order to identify useful fairness based trends
- NLP_Fairness_in_stories.ipynb
- Fetches the data + Preprocessing to output clean sentences from stories corpus
- Extracts Events using Spacy
- Categorizes events on gender and filters out non-gender based events
- Visualizations.ipynb
- Visualizes odds ratio scores of our extracted events
- weat.ipynb
- Computes the weat scores using the word2vec model for the 5 different genre-based categories of our stories dataset:
- Genre-based Categories: General, Children, Fiction, Male_Author, Female_Author
Active internet connection is required for downloading stories from Gutenberg (we are using html links to fetch stories)