부산대학교 도서관

A power of an online user can start on a single click. It is now easier to spread news on cyberspace. This can be done on social media or any other websites online. But not everything online is real. Fake news emerges with a malicious intent that might harm and cause trouble. Primarily online wherein it is easy to publish and spread false information in seconds. This paper presents the development of a fake news detection on English news article's title using Machine learning algorithm as fake news continuously spread worldwide. With the use of ensembling method, Naïve Bayes achieved 54%. Similarly, ensembling technique is applied to another algorithm, XGBoost that garnered 80%. Furthermore, SVM was also added on feature combinations that are also explored to further understand which feature combination can give a better performance. Findings of the study shows that the ensemble learning techniques can be used to identify the accuracy of fake news in title alone.