This is my Machine Learning model created with PassiveAggressiveClassifier to detect a news as Real or Fake depending on it's contents. It uses TfidfVectorizer which converts a collection of raw documents into a matrix of TF-IDF features.
Upon running the Jupyter Notebook we get the accuracy of 93.05% which is not that great but reasonable enough!
Dataset - https://drive.google.com/file/d/1er9NJTLUA3qnRuyhfzuN0XUsoIC4a-_q/view
This is a small ML project created just for learning purposes.