Claim Missing Document
Check
Articles

Found 1 Documents
Search

Klasifikasi Hoaks Berbahasa Inggris menggunakan Boosting Weighted Extreme Learning Machine Luthfi Mahendra; Indriati Indriati; Putra Pandu Adikara
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 3 No 8 (2019): Agustus 2019
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1210.53 KB)

Abstract

Rapid technological developments have caused hoax to be more easily disseminated through the internet, especially for politic related news. Although it looks trivial, hoax can cause various kinds of problems such as community riots and the blocking of social media sites. To overcome the problems that can be caused by hoaxes, this study attempts to create an automatic English language hoax classification system using the Weighted Boosting ELM algorithm. The algorithm was chosen because it has high accuracy results for various types of document classification problems and has good results even if the data used has an unbalanced number of classes, making it suitable for hoax classifications which are fewer than factual news. The research methodology is divided into several stages consisting of pre-processing, term weighting, normalization, training and algorithm evaluation. The data used are 180 articles consisting of 90 hoax and 90 factual news. Evaluation was carried out by measuring F1 values ​​(results of average harmonic precision and recall) using K-Fold cross validation, the highest results obtained were 0,787.