Skip to content

RehanShaikh-ai/jigsaw-text-moderation-with-xai

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Jigsaw Toxic Comment Classification

License Status Python Data


Overview

Jigsaw Toxic Comment Classification leverages the Jigsaw dataset to build and benchmark text moderation systems.
The aim is to compare traditional machine learning models with modern deep learning and explore explainable AI (XAI) for trustworthy moderation.

Model XAI Build


Features

  • Data preprocessing of the Jigsaw public dataset
  • Baseline ML models (e.g., Logistic Regression, SVM)
  • Enhanced Deep Learning model (possible LSTM, CNN, Transformers)
  • Explainable AI integration (XAI) for model transparency
  • Clean and documented codebase for reproducibility

Notebooks Pull Requests


Tech Stack

  • Python 3.8+
  • Pandas, scikit-learn, NumPy for ML pipeline
  • TensorFlow/PyTorch, Keras for deep learning models
  • LIME/SHAP for Explainable AI (XAI)
  • Notebooks for EDA and experiments

Dataset

Jigsaw Toxic Comment Classification Challenge dataset.
More details can be found on Kaggle.

Dataset Size

Getting Started

Setup instructions and sample usage will be added soon, once the pipeline is finalized.


Usage

Examples and details on training, evaluation, and inference will be provided after completing initial model implementation.


Contributing

  • Please open an issue to suggest features or report bugs.
  • Pull requests are welcome.
  • All contributors are expected to follow the code of conduct.

Contributors Discussions


License

MIT


👤 Author

Rehan Abdul Gani Shaikh
Aspiring Data Scientist | B.Tech Student

🔗 Connect with me: LinkedIn
📬 Email: rehansk.3107@gmail.com


README will be updated to reflect setup, data usage, and model details as the project evolves.

About

Any suggestions, improvements and corrections are appreciated !

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published