Enhancing False News Detection through Supervised Machine Learning and NLP Techniques: A Comparative Study of Feature Extraction and Selection Methods Using Python Scikit-Learn

The spread of false news via social and other media platforms is a major worry because it has the potential to cause significant harm to societies and nations. As a result, many academics are trying hard to detect and counteract the spread of fake news. This study analyses research on false news identification and provides a novel strategy that uses supervised machine learning algorithms, standard machine learning models, and textual analysis tools such as Python, Scikit-Learn, and NLP. The suggested method effectively classifies news articles as true or fake by utilizing techniques such as feature extraction and vectorization. For text data tokenization and feature extraction, the Python Scikit- Learn module, known for useful functions such as Count Vectorizer and Tiff Vectorizer, is used. Furthermore, feature selection methods are used to experiment with and discover the most appropriate features, resulting in maximum precision based on the confusion matrix analysis. The proposed model is trained on a dataset that includes both true and false news articles, allowing it to classify new articles quickly and accurately, assisting users in establishing the legitimacy of online news stories. In summary, this study contributes to continuing efforts to combat false news by reviewing previous research on its detection and providing an approach that combines classical machine learning models, Python Scikit-Learn, and natural language processing (NLP) for effective textual analysis. The suggested model detects false news with high accuracy and precision, providing users with a useful tool for determining the validity of news pieces.