Analysis of Indonesian Netizen Sentiment on Platform X Regarding the Arrival of Refugees in Indonesia Using the Multinominal Naive Bayes Method

Abstract

This research aims to analyze the sentiments of Indonesian netizens regarding the arrival of Rohingya refugees in Indonesia using the Multinomial Naive Bayes method. Sentiment analysis was carried out on comments obtained from platform X. The data collection technique used the crawling method to extract comments from platform X users regarding the issue of the arrival of Rohingya refugees. The tool used for crawling is Google Collab. The data analysis process includes sentiment labeling, data preprocessing (case folding, stopword removal, tokenizing, stemming), and classification using the Multinomial Naive Bayes method. The research results show that the majority of Indonesian netizens' sentiments regarding the arrival of Rohingya refugees in Indonesia are negative, with a percentage of 81%. Positive sentiment reached 8%, while neutral sentiment was 11%. The Multinomial Naive Bayes method produces an accuracy of 82.5% in classifying netizen sentiment. The tools used to process the data are the Orange Data Mining application version 3.36.2 It is hoped that this research can contribute to the development of computer science, especially in the fields of Text Mining, Natural Language Processing, Machine Learning and Artificial Intelligence (AI). It is also hoped that this research will provide benefits to parties related to handling the Rohingya refugee problem in Indonesia, such as the government, humanitarian organizations, mass media, academics, the general public, and other researchers.