A Deep Learning approach of classifying the news headlines and content as Fake or Real.
A considerable amount of labeled news headlines and content are taken and a Deep Learning approach is used to classify any news as Fake or Real. Text pre-processing is done using NLTK library. The words are converted into vectors using Word Embeddings. The model is built using LSTM and Bi-directional LSTM with Dropout Layers.
It was found that LSTM out-performed Bi-diectional LSTM for this use-case.