Analisis Sentimen Konten Radikal Di Media Sosial Twitter Menggunakan Metode Support Vector Machine (SVM)

Ferdi Alvianda (Fakultas Ilmu Komputer, Universitas Brawijaya)
Indriati Indriati (Fakultas Ilmu Komputer, Universitas Brawijaya)
Putra Pandu Adikara (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 3 No 1 (2019): Januari 2019

Lately, there are many terrorist threats by the radicals in Indonesia. Radicals keep growing by numbers each day as they share their radical beliefs to other people. These radical beliefs can be shared through social media, such as Twitter. Therefore, a research regarding that problem is conducted. Documents of Twitter that contain radical tweets are classified to two categories, positive radical content and negative radical content. The method used for this research is Support Vector Machine (SVM) with Polynomial Degree Kernel. The highest accuracy rate achieved from this research is 70% with the parameter value of λ is 0,1, constant value of γ is 0,1, maximum iteration of 5 with training data sets of 80 documents (60 negative documents and 20 positive documents) as training data sets and 20 documents (15 negative documents and 5 positive documents) as testing data sets.