In this project I implemented neural networks to detect fake news articles. I did this by two methods:
The first one uses TF-IDF to classify and detect fake news. This was approximately 92% accurate. I used this model as a reference to create my neural network: https://data-flair.training/blogs/advanced-python-project-detecting-fake-news/
The dataset I used for this model is in this repository by the name of news.zip. It has about 8000 articles that are either real or fake.
The second one uses LSTM cells and Recurring Neural Network (RNN) to detect and classify fake news. This model was nearly 99% accurate which is significantly more accurate than the first method.
The dataset I used for this model is on kaggle: https://www.kaggle.com/clmentbisaillon/fake-and-real-news-dataset. It consists of about 40000 articles that are either real or fake news. I used this model as a reference: https://medium.com/analytics-vidhya/fake-news-detection-with-lstm-a0f7aeeca982