Analisis Sentimen Cuitan X terhadap Pendaki Asing Gunung Rinjani Menggunakan Algoritma SVM dan Random Forest

Abstract

Social media platforms have become significant sources of public opinion regarding various issues, including tourism activities. This study analyzes public sentiment towards foreign climbers on Mount Rinjani through X (Twitter) social media posts using Support Vector Machine (SVM) and Random Forest algorithms. Data collection employed web scraping techniques with six relevant keywords, resulting in 4,777 unique tweets after cleaning and duplicate removal. The lexicon-based labeling approach revealed an imbalanced distribution with 63.58% negative sentiment, 18.8% positive sentiment, and 17.63% neutral sentiment. To address class imbalance, Synthetic Minority Oversampling Technique (SMOTE) was applied, creating a balanced dataset of 9,111 samples. Text preprocessing included case folding, normalization, tokenization, stopword removal, and stemming using Indonesian language tools. Feature extraction utilized TF-IDF vectorization with parameters optimized for Indonesian text analysis. The dataset was split into 70% training, 15% validation, and 15% testing using stratified sampling. Evaluation results demonstrated that SVM achieved superior performance with 95.7% accuracy, 96% precision, 95.7% recall, and 95.7% F1-score, while Random Forest achieved 94.4% accuracy, 94.4% precision, 94.4% recall, and 94.4% F1-score. The dominance of negative sentiment indicates public concerns regarding foreign climbing activities that require stakeholder attention. This research contributes to sentiment analysis methodology for Indonesian social media text and provides practical insights for sustainable tourism management in Mount Rinjani National Park