Fakultas Teknologi Informasi

Implementation of Fake News Detection Using Long Short Term Memory Method Base on Android

Penulis
Dosen:
  1. TONI ARIFIN
Eksternal:
  1. Gunawansyah
  2. Faris Ghilmany
  3. Riffa Haviani Laluma
  4. Bambang Sugiarto
Tanggal Terbit
07 September 2022
Kategori
Seminar Internasional [Lainnya]
Penerbit
International Conference on Telecomunication System, Services, and Aplication (TSSA)
Kota / Negara
Bali / Indonesia
Halaman
1-5
E-ISBN
10.1109/TSSA52866.2021.976828
URL
https://ieeexplore.ieee.org/document/9768283/
Abstrak
Technology is a part of industrial revolution that can make human easier to receipt informations and news. In 2019, 63,3% Indonesian people already have personal smartphone and owning a smartphone can indicate an accelerator to getting news. But it is an opportonity for irresponsible people to share fake news. Therefore it is necessary to make a system to help people in detecting fake news, include in smartphone application. In this research, a machine learning model was made using the LSTM (LongShort Term Memory) method to detect fake news. LSTM is a classification method that improves the weakness of the previous method, namely RNN (Recurrent Neural Network). An android application was built in this study to make it easier to detect fake news. An application programming interface was also built to bridge the detection process on the backend to the Android application. The results of this study indicate that the LSTM method is able to detect fake news with an accuracy value of 0.983. The evaluation results of the model training, the model does not show the occurrence of underfitting or overfitting.