NikosMav/FakeNews-Classification ? reverse-engineered prompt

Build me a Jupyter notebook project that classifies news articles as fake or real using machine learning.

I want it to load a news dataset, clean and prepare the text, then compare a few common approaches for turning text into numbers, like simple word counts, TF IDF, and word embeddings. After that, train and test a few models such as logistic regression, naive Bayes, support vector machine, and random forest, then show which one performs best using accuracy and F1 score.

Please make it easy to follow, with clear steps and comments, and include basic charts or summaries so I can see how the models compare. If you need to look up the latest library usage or best practices, go ahead and check the current docs online. I’d also like the notebook to include an improved version of the random forest setup if that helps boost results.