Fake News Classification Using SVM and Logistic Regression Methods

Digital media has become a platform for people to obtain information quickly in this digital age. Through digital media, fake news can be quickly and easily spread to the public, causing anxiety, conflict and harm to the public. Therefore, there is a need to detect and classify hoax news accurately. This research aims to create a classification model using SVM and Logistic Regression methods. The dataset used in this research consists of 5932 news, with 2893 fake news and 3039 real news. The dataset goes through preprocessing, TF-IDF transformation, training the Logistic Regression model to find the initial classification probability, and training the linear kernel SVM model to find the final classification. The model was then tested on the tested data. The results showed that the model has an accuracy of 96.4%, a precision of 96.7<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, a sensitivity of 95.7<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, and an F1-score range of 96.2%. These results show that the SVM and Logistic Regression methods are effective in classifying fake news.