Toxic Comment Classification challenge from Kaggle. Used for CS 7650 final project at Georgia Tech, Spring 2021.
From https://www.kaggle.com/c/jigsaw-toxic-comment-classification-challenge/overview
Steps:
- Download data and gloVe vectors (see READMEs in corresponding folders for links)
- Run preprocessing notebooks
- Clean data
- Create vocab/filter gloVe
- Convert comments (words) to numbers (indices) used in neural networks
- Models
- Run respective jupyter notebook for model you wish to build.
Notes: - Notebooks were built in Google Colab, utilizing free GPUs.