Fake News Detection Using Machine Learning

Abstract

Fake news on social media and other platforms is widespread, and it is a reason for great concern because of its potential to inflict significant social and national harm with negative consequences. Detection is already the topic of a lot of studies. This paper is a good example of news detection. The study on fake news identification is examined, as well as the traditional machine learning methods. Selective learning models to select the best ones in order to construct a product model with supervised learning. Using technologies like Python, a machine learning system can classify fake news as true or false. Use NLP for textual analysis with Scikit-learn. As a result of this procedure, features will be extracted and vectorized. To do tokenization and feature extraction, we recommend utilizing the Python scikit-learn module. Because this library offers important functions like count vectorizer and tiff, text data can be extracted. Then we'll experiment with feature selection approaches to find the best one. According to the confusion matrix results, fit features to acquire the highest precision. I use some machine learning algorithms techniques to detect the fake news. Those are the multinomial NB, Naive Bayes and BernoulliNB, and logistic Regression. Nevertheless, I did uncover promising setups for both purposes. I got the best accuracy from Logistic Regression which was 89%.