Fake News Detection
Compared BERT, CNN, and LSTM architectures for fake news detection. Fine-tuned BERT achieved the highest F1-score on labeled news corpora, outperforming traditional NLP baselines.
Technologies
- Python
- TensorFlow
- Keras
- Plotly
- NetworkX
- Pandas
- NumPy
- Scikit-learn
Main features
- Comparison of BERT, CNN, and LSTM classifiers on labeled news corpora
- Fine-tuned BERT achieving the highest F1-score of the three models
- Text preprocessing and feature pipelines with pandas and scikit-learn
- Result and relationship analysis visualized with Plotly and NetworkX