Classification of the likelihood of Indonesian Facebook users in spreading hoaxes using Support Vector Machine (SVM)

T. V. Rampisela, H. T. Andarlia, Z. Rustam

Research output: Contribution to journalConference articlepeer-review

Abstract

Social media is the most commonly accessed internet content by Indonesian netizens; 97.4 % Indonesians access social media while using the internet. In 2016, Facebook was the most commonly used social media format for Indonesian internet users. While many are benefited by the features offered by Facebook, many also use Facebook for things that they are not supposed to, such as sharing hoaxes, either intentionally or unintentionally. This clearly puts Facebook users at a disadvantage, considering the increasing trend of hoax-sharing nowadays. Therefore, this research aims to prevent the spread of hoax through Facebook by analyzing the pattern of how people use Facebook. This pattern is obtained from a survey on 200 samples that use Facebook, chosen by purposive sampling. Using the Support Vector Machine method, an application of the collaboration between mathematics and computer science, the acquired data is used to predict whether or not someone has the potential of spreading hoaxes. Simulation results show that the average of the prediction accuracy of this binary classification problem is 86 percent. Hence, it is hoped that Facebook could prevent the sharing of hoaxes by making use of the results from this research.

Original languageEnglish
Article number012019
JournalJournal of Physics: Conference Series
Volume1725
Issue number1
DOIs
Publication statusPublished - 12 Jan 2021
Event2nd Basic and Applied Sciences Interdisciplinary Conference 2018, BASIC 2018 - Depok, Indonesia
Duration: 3 Aug 20184 Aug 2018

Keywords

  • Classification
  • Facebook
  • Hoax
  • Support vector machines

Fingerprint Dive into the research topics of 'Classification of the likelihood of Indonesian Facebook users in spreading hoaxes using Support Vector Machine (SVM)'. Together they form a unique fingerprint.

Cite this