akshayjadhav116/ai-fake-news-detector ? reverse-engineered prompt

Build me a simple Streamlit web app that can tell whether a news article looks fake or real.

I want a clean page where someone can paste in a full article, load a sample article, or clear the text and try again. When they click analyze, it should preprocess the text, turn it into features with TF IDF, and use a trained machine learning model to make a prediction. Show the result clearly, along with a confidence style display, the active model name, the model accuracy, and the dataset size on the page.

Please make it feel polished and easy to use, with a small info section that explains what the app does and a note that the prediction is only a helpful signal, not final proof. If there are saved model files and a vectorizer available, use those. If you need to look up current Streamlit docs online, go ahead.