LSTM Fake News Detection System
A deep-learning text classifier that distinguishes real and fake news using a stacked LSTM architecture and a documented evaluation pipeline.
Status:Portfolio project; repository reports 93.9% test accuracy
View source on GitHub (opens in a new tab)- Services
- Machine Learning
- Stack
- Python, Pandas, TensorFlow, Keras +3 more
Outcomes
The results
93.9%
Reported test accuracy
01
Our solution
The project prepares text with NLP tooling, trains a stacked LSTM model in TensorFlow and Keras, and evaluates predictions with accuracy, precision, recall, and F1 score.
02
The outcome
Demonstrates an end-to-end deep-learning classification pipeline and clear reporting of model performance.
03
Best for
Content credibility research, moderation experiments, NLP classification, and deep-learning portfolio work.
Technology
The stack behind it
- Backend
- Python
- Data & BI
- Pandas
- NumPy
- AI & ML
- TensorFlow
- Keras
- LSTM
- NLTK