Implementasi Algoritma K-NN Pada Sosial Media X Untuk Analisis Sentimen Pengalaman Warganet Tinggal Di Luar Negeri

Abstract

The development of information technology, especially through social media such as Twitter, has changed the way people search for information. With more than 6.43 million users in Indonesia, Twitter has become the main platform for sharing opinions. This study aims to analyze the sentiments of Indonesian citizens (WNI) living abroad, who often face challenges and opportunities in adapting to new environments. Given the increasing number of WNI, reaching over 9 million in 2020, understanding their sentiments is crucial. The K-Nearest Neighbor (KNN) method was used to classify sentiments as positive, negative, or neutral. This study involved data collection through the tweet-harvest technique, where 1,060 comments were successfully collected, and 600 of them met the relevance criteria for analysis. The analysis results showed that 60.4% of sentiments were neutral, 34.1% were positive, and 5.5% were negative, with the KNN model achieving an accuracy of 81.67%. Model evaluation revealed the highest precision in the neutral class and a recall of 1.00, although the positive and negative classes require further optimization. This study is expected to provide insights for the public and decision-makers regarding the experiences of Indonesian citizens abroad.