Sentiment Analysis on Indonesian National Football Team Naturalization using KNN and SVM

Abstract

The naturalization of football players in Indonesia is largely viewed positively, with supporters highlighting its benefits for team performance, international competitiveness, and player development. While PSSI endorses naturalization to strengthen the national team, Liga Indonesia Baru (PT LIB) imposes limits to maintain fairness. The purpose of this research is to examine public sentiment toward the naturalization of Indonesian football players by analysing discussions on X and YouTube. This research analyses public sentiment toward the naturalization of Indonesian football players using a data and text mining approach based on 3,267 comments from X and YouTube between 2022 and 2024. The research process includes data collection, preprocessing, TFIDF, data labeling, and model training and evaluation. Two machine learning models, KNN and SVM, are implemented for classification, with SVM outperforming KNN in accuracy. Our results show that KNN achieved 76.71% accuracy (precision: 52%, recall: 56%, F1-score: 53%), while SVM RBF outperformed with 86.51% accuracy (precision: 59%, recall: 42%, F1-score: 26%). SMOTE and GridSearch effectively address the class imbalance and optimize model performance. Public sentiment is predominantly positive, highlighting enhanced team performance and global recognition. These insights assist PSSI and policymakers in making informed decisions regarding fairness, discrimination, and the governance of Indonesian football.