for data set please follow this link: https://www.kaggle.com/c/fake-news/data
It's a NLP project solved with machine learning module called MultinomialNB steps followed:
- Load the data
- Import the required libraries
- preprocessing Data
- cleaning the null values
- removing the unwanted text, dots etc.
- Vectorizing the cleaned data with TF-IDF technique
- training and Testing the model
- Result with Accuracy of 91% (MultinomialNB) and 99% (LSTM)
Multinominal model accuracy is by taking all columns but LSTM is only for title and author columns beacuse while taking all the columns the time taken is more.