Analisis Sentimen Video YouTube KOMPASTV “Pajak Cekik Rakyat, Tunjangan DPR Naik” Menggunakan Naive Bayes & SVM

Abstract

This study aims to analyze public sentiment regarding flooding in Sumatra based on data from social media platform X. Flooding is a frequent natural disaster in Sumatra and elicits a variety of public responses, many of which are expressed through social media. Social media platform X was chosen as the data source because it is open and real-time, allowing it to broadly represent public opinion.

The research data consists of 1,030 Indonesian-language tweets collected through a crawling process using the official API for X, using keywords related to flooding in Sumatra. After data cleaning, 873 tweets were obtained, which were then processed through text mining stages, including text preprocessing, manual sentiment labeling, and dividing the data into training and test data. The training data consisted of 650 tweets, while the test data consisted of 223 tweets.

Sentiment classification was performed using the Naive Bayes algorithm with the assistance of RapidMiner software. Model evaluation was performed using a confusion matrix with accuracy, precision, recall, and F1-score metrics. The results showed that the Naive Bayes algorithm performed quite well in sentiment classification. Furthermore, the analysis shows that public opinion regarding the flooding in Sumatra is dominated by negative sentiment. This research is expected to provide insight into public perceptions and inform disaster management policymaking.