- This repository contains the implementation and experiments conducted on building a scalable misinformation detection system using DistilBERT and multiple real-world fake news datasets.
- The project aims to detect misinformation by leveraging textual content (news body + headline) across different social media news sources while maintaining efficiency and scalability using distilled transformer models.
- Built a lightweight and optimized deep learning model using transformer-based encoders and PyTorch, achieving 80% accuracy while reducing computational load for deployment in low-resource environments.
- Fine tuned BERT for better results.
i-wav/MisinformationDetection
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|