PERBANDINGAN METODE KLASIFIKASI NAÏVE BAYES, DECISION TREE, RANDOM FOREST TERHADAP ANALISIS SENTIMEN KENAIKAN BIAYA HAJI 2023 PADA MEDIA SOSIAL YOUTUBE

Abstract

Euphoria The implementation of the 2023 Hajj pilgrimage has become a trending topic of conversation on social media, one of which is on YouTube. The government of the Kingdom of Saudi Arabia implemented a policy to reduce the cost of Hajj packages by 30% from the previous year and there was no age limit for pilgrims and even received an additional quota of 221,000 pilgrims for Indonesia. Contrary to the policy of the Saudi Arabian government, the Ministry of Religion created a contradictory discourse by proposing an increase in the costs of the Hajj with the composition of the financing being 70% of the costs of the Hajj trip (BIPIH) and 30% of the value of the benefits in the proposed Hajj costs in 2023. The proposed Hajj costs were ratified become a policy through Presidential Decree Number 7 of 2023 regarding the costs of organizing the pilgrimage in 1444 hijriah/2023. This gave rise to various public responses to the policy. The purpose of this study was to analyze public sentiment by classifying positive and negative comments on YouTube social media regarding the policy of increasing the cost of the 2023 Hajj pilgrimage. This research compared the results of the accuracy of several classification methods such as naïve bayes, decision trees and random forest. From this research, the accuracy of naïve Bayes was obtained by 90%, Decision Tree by 83%, Random Forest by 87%. It is hoped that with the sentiment analysis of the classification of YouTube comments, it can provide an overview for the government to see the response of the public's assessment in making a policy.