Analisis Sentimen Opini Publik terhadap Fenomena Bunuh Diri Mahasiswa di Twitter Menggunakan Algoritma Naive Bayes

Abstract

Currently, Twitter or X social media has become a popular social media for interacting and providing opinions on trending events. The opinion that is still widely discussed is the phenomenon of student suicide. With the many opinions on Twitter social media about the phenomenon of student suicide, this creates a dilemma about the phenomenon of student suicide, whether student suicide is a positive or negative behavior. For this reason, sentiment analysis was carried out to group opinions on Twitter social media. In analyzing this sentiment, the Naive Bayes algorithm is used, which is a data mining algorithm that has the advantage of being easy in the calculation stage. The rapidminer application is used to analyze opinions on Twitter social media. The results obtained by the Naive Bayes algorithm provide an accuracy of 60%. Sentiment analysis shows that there are 41% of opinions that have positive sentiment values ​​and there are 59% of opinions that have negative sentiment values. So the majority of opinions about student suicide on Twitter social media have negative sentiment values. In collecting opinion data, it also shows that the words news people and semarang are words that are often written on Twitter social media.