Sentiment Analysis on the FIFA U-20 World Cup in Argentina Using Support Vector Machine

Authors

  • Achmad Warsito Sujatmiko Informatics Department, Universitas Dr. Soetomo, Surabaya
  • Anik Vega Vitianingsih Informatics Department, Universitas Dr. Soetomo
  • Slamet Kacung Informatics Department, Universitas Dr. Soetomo, Surabaya
  • Dwi Cahyono Informatics Department, Universitas Dr. Soetomo, Surabaya
  • Anastasia Lidya Maukar Industrial Engineering Department, President University, Bekasi

Keywords:

FIFA World Cup U-20 Argentina, Soundtrack World Cup U-20 2023, Sentiment Analysis, Support Vector Machine

Abstract

The decision made by FIFA regarding the selection of the soundtrack and the host country for the FIFA U-20 World Cup has sparked emotional reactions among the public and raised concerns about the event, especially on social media platform X. This is due to FIFA’s decision to choose a soundtrack not from the host country, Argentina, but from the previous host, Indonesia. FIFA should advocate for the creation of a soundtrack by the host country to reflect its distinctive characteristics or atmosphere. Concerns about the U-20 World Cup in Argentina have also been fueled by the country’s economic crisis, which is feared to affect the facilities and infrastructure for the young players representing their nations. This research focuses on filtering public responses to FIFA’s decisions regarding the soundtrack selection and the host country for the U-20 World Cup into positive, neutral, and negative categories using the Support Vector Machine (SVM) method. The research aims to provide policy recommendations regarding the host selection process and cultural representation in international sports events. Additionally, this study is expected to provide a deeper understanding of the preferences and values held by the public regarding international sports. The research steps include data collection, pre-processing, labeling, weighting, and classification using a Support Vector Machine. The data for this research were obtained through crawling on social media platform X, totaling 2400 data points. The performance evaluation of the SVM algorithm using a 50:50 ratio of training and testing data yielded an average accuracy of 85.71%, Precision of 85.98%, Recall of 85.71%, and F1-score of 85.58%.

Published

15-06-2024