CD Skripsi
Verifikasi Berita Palsu Pada Twitter Berbasis Deep Learning Menggunakan Algoritma Rnn
ABSTRACT
Twitter has become a social media that its users use to find the latest news. Many news are spread on twitter social media, making users have to be wise in looking at the information that is spread. It was found that fake news on social media twitter was retweeted 70% more than the original news. Fake news is very dangerous if the public cannot filter it properly. Fake news can disturb the community and even threaten the integrity of the nation. It is therefore important to stop the spread of fake news. This research developed a technique that is able to verify fake news on Twitter based on Deep Learning using the Recurrent Neural Network (RNN) algorithm. The dataset comes from the results of Crawling Twitte in Indonesian as many as 1700 data for the 2019-2022 period. The results of the research from the creation of a fake news verification model on Twitter based on deep learning using the RNN algorithm produced a model that had an accuracy of more than 90% on the epoch parameter of 30, and a batch size of 15 with a model accuracy value of 92.5% using a confusion matrix.
Keywords: fake news, twitter, Recurrent Neural Network (RNN)
Tidak tersedia versi lain