<![CDATA[The development of information technology allows people to access news anywhere and anytime through the internet and social media. However online news often contains facts that are manipulated to mislead readers, which is called fake news (hoax). Fake news or hoaxes are information that is created and manipulated with the aim of misleading readers. Fake news is generally created with provocative and slanderous titles, with language that can convince readers. Fake news can spread quickly due to people's ignorance in filtering the information they get, and unconsciously spreading fake news to others. Therefore, it is necessary to create a system that can help people to distinguish between fake news and news that is fact. This research aims to create a fake news detection system using the Long Short Term Memory (LSTM) algorithm using Indobert. This research uses a dataset containing 3309 hoax and non-hoax news. This model can predict fake news with an accuracy rate of 94.2%, precision 0.944, recall 0.940 and f1 score of 0.941 which is better than the comparison method. Keywords: hoax, LSTM, hoax prediction system, deep learning, Indobert]]>
Copyrights © 2023
