In a world of negativity and ruse, what's worse is to not do anything about rumors and headlines that are just well-calculated ways to capture clout and earn tentative fame. Gaining inspiration from krishnaik06, I designed a Fake News Classifier using the following concepts:
- Word Embedding: in calculating the semantics association between words by converting the textual data to numeral vectors
- One-Hot Encoding: in transforming the categorical attributes into binary vectors which acts as an input to the LSTM model
- LSTM: in finally classifying the news into fake or not by processing the encoded data
Please find the ipnyb files for each attached to this project merely for brushing up on concepts.
I have chosen not to make the code of this project publicly available because the development process closely followed the methodology demonstrated by Krish Naik. As this project was primarily an experiment and closely mirrored the work of Mr. Naik, I believe that making it publicly accessible would amount to replicating his efforts. I respect his work and do not intend to present a derivative of it as an original creation.